Research

@article {Hu2024.11.17.623906,
  author = {Hu, Bo and Temiz, Nesibe Z. and Chou, Chi-Ning and Rupprecht, Peter and Meissner-Bernard, Claire and Titze, Benjamin and Chung, SueYeon and Friedrich, Rainer W},
  title = {Representational learning by optimization of neural manifolds in an olfactory memory network},
	elocation-id = {2024.11.17.623906},
	year = {2024},
	doi = {10.1101/2024.11.17.623906},
  publisher = {Cold Spring Harbor Laboratory},
  journal = {bioRxiv}
}

Higher brain functions depend on experience-dependent representations of relevant information that may be organized by attractor dynamics or by geometrical modifications of continuous "neural manifolds". To explore these scenarios we analyzed odor-evoked activity in telencephalic area pDp of juvenile and adult zebrafish, the homolog of piriform cortex. No obvious signatures of attractor dynamics were detected. Rather, olfactory discrimination training selectively enhanced the separation of neural manifolds representing task-relevant odors from other representations, consistent with predictions of autoassociative network models endowed with precise synaptic balance. Analytical approaches using the framework of manifold capacity revealed multiple geometrical modifications of representational manifolds that supported the classification of task-relevant sensory information. Manifold capacity predicted odor discrimination across individuals, indicating a close link between manifold geometry and behavior. Hence, pDp and possibly related recurrent networks store information in the geometry of representational manifolds, resulting in joint sensory and semantic maps that may support distributed learning processes.Competing Interest StatementThe authors have declared no competing interest.

@article{MCC24,
  title={Nonlinear classification of neural manifolds with contextual information},
  author={Francesca Mignacco and Chi-Ning Chou and SueYeon Chung},
  year={2024},
  eprint={2405.06851},
  archivePrefix={arXiv},
  primaryClass={q-bio.NC}
}

Understanding how neural systems efficiently process information through distributed representations is a fundamental challenge at the interface of neuroscience and machine learning. Recent approaches analyze the statistical and geometrical attributes of neural representations as population-level mechanistic descriptors of task implementation. In particular, manifold capacity has emerged as a promising framework linking population geometry to the separability of neural manifolds. However, this metric has been limited to linear readouts. Here, we propose a theoretical framework that overcomes this limitation by leveraging contextual input information. We derive an exact formula for the context-dependent capacity that depends on manifold geometry and context correlations, and validate it on synthetic and real data. Our framework's increased expressivity captures representation untanglement in deep networks at early stages of the layer hierarchy, previously inaccessible to analysis. As context-dependent nonlinearity is ubiquitous in neural systems, our data-driven and theoretically grounded approach promises to elucidate context-dependent computation across scales, datasets, and models.

@article {Chou2024.02.26.582157,
  author = {Chi-Ning Chou and Luke Arend and Albert J Wakhloo and Royoung Kim and Will Slatton and SueYeon Chung},
  title = {Neural Manifold Capacity Captures Representation Geometry, Correlations, and Task-Efficiency Across Species and Behaviors},
  elocation-id = {2024.02.26.582157},
  year = {2024},
  doi = {10.1101/2024.02.26.582157},
  publisher = {Cold Spring Harbor Laboratory},
  journal = {bioRxiv}
}

The study of the brain encompasses multiple scales, including temporal, spatial, and functional aspects. To integrate understanding across these different levels and modalities, it requires developing quantification methods and frameworks. Here, we present effective Geometric measures from Correlated Manifold Capacity theory (GCMC) for probing the functional structure in neural representations. We utilize a statistical physics approach to establish analytical connections between neural co-variabilities and downstream read-out efficiency. These effective geometric measures capture both stimulus-driven and behavior-driven structures in neural population activities, while extracting computationally-relevant information from neural data into intuitive and interpretable analysis descriptors. We apply GCMC to a diverse collection of datasets with different recording methods, various model organisms, and multiple task modalities. Specifically, we demonstrate that GCMC enables a wide range of multi-scale data analysis. This includes quantifying the spatial progression of encoding efficiency across brain regions, revealing the temporal dynamics of task-relevant manifold geometry in information processing, and characterizing variances as well as invariances in neural representations throughout learning. Lastly, the effective manifold geometric measures may be viewed as order parameters for phases related to computational efficiency, facilitating data-driven hypothesis generation and latent embedding.

@article{PRXQuantum.5.010334,
  title = {Limitations of Linear Cross-Entropy as a Measure for Quantum Advantage},
  author = {Gao, Xun and Kalinowski, Marcin and Chou, Chi-Ning and Lukin, Mikhail D. and Barak, Boaz and Choi, Soonwon},
  journal = {PRX Quantum},
  volume = {5},
  issue = {1},
  pages = {010334},
  numpages = {27},
  year = {2024},
  month = {Feb},
  publisher = {American Physical Society},
  doi = {10.1103/PRXQuantum.5.010334},
  url = {https://link.aps.org/doi/10.1103/PRXQuantum.5.010334}
}

Demonstrating quantum advantage requires experimental implementation of a computational task that is hard to achieve using state-of-the-art classical systems. One approach is to perform sampling from a probability distribution associated with a class of highly entangled many-body wavefunctions. It has been suggested that this approach can be certified with the Linear Cross-Entropy Benchmark (XEB). We critically examine this notion. First, in a "benign" setting where an honest implementation of noisy quantum circuits is assumed, we characterize the conditions under which the XEB approximates the fidelity. Second, in an "adversarial" setting where all possible classical algorithms are considered for comparison, we show that achieving relatively high XEB values does not imply faithful simulation of quantum dynamics. We present an efficient classical algorithm that, with 1 GPU within 2s, yields high XEB values, namely 2-12% of those obtained in experiments. By identifying and exploiting several vulnerabilities of the XEB, we achieve high XEB values without full simulation of quantum circuits. Remarkably, our algorithm features better scaling with the system size than noisy quantum devices for commonly studied random circuit ensembles. To quantitatively explain the success of our algorithm and the limitations of the XEB, we use a theoretical framework in which the average XEB and fidelity are mapped to statistical models. We illustrate the relation between the XEB and the fidelity for quantum circuits in various architectures, with different gate choices, and in the presence of noise. Our results show that XEB's utility as a proxy for fidelity hinges on several conditions, which must be checked in the benign setting but cannot be assumed in the adversarial setting. Thus, the XEB alone has limited utility as a benchmark for quantum advantage. We discuss ways to overcome these limitations.

@journal{CGSV24J,
  author = {Chou, Chi-Ning and Golovnev, Alexander and Sudan, Madhu and Velusamy, Santhoshini},
  title = {Sketching Approximability of All Finite CSPs},
  year = {2024},
  publisher = {Association for Computing Machinery},
  address = {New York, NY, USA},
  issn = {0004-5411},
  url = {https://doi.org/10.1145/3649435},
  doi = {10.1145/3649435},
  journal = {J. ACM},
}

A constraint satisfaction problem (CSP), $\textsf{Max-CSP}(\mathcal{F})$, is specified by a finite set of constraints $\mathcal{F} \subseteq \{[q]^k \to \{0,1\}\}$ for positive integers $q$ and $k$. An instance of the problem on $n$ variables is given by $m$ applications of constraints from $\mathcal{F}$ to subsequences of the $n$ variables, and the goal is to find an assignment to the variables that satisfies the maximum number of constraints. In the $(\gamma,\beta)$-approximation version of the problem for parameters $0 \leq \beta < \gamma \leq 1$, the goal is to distinguish instances where at least $\gamma$ fraction of the constraints can be satisfied from instances where at most $\beta$ fraction of the constraints can be satisfied. In this work we consider the approximability of this problem in the context of sketching algorithms and give a dichotomy result. Specifically, for every family $\mathcal{F}$ and every $\beta < \gamma$, we show that either a linear sketching algorithm solves the problem in polylogarithmic space, or the problem is not solvable by any sketching algorithm in $o(\sqrt{n})$ space.

@inproceedings{KCPDDSC24,
title={Probing Biological and Artificial Neural Networks with Task-dependent Neural Manifolds},
author={Michael Kuoch and Chi-Ning Chou and Nikhil Parthasarathy and Joel Dapello and James J. DiCarlo and Haim Sompolinsky and SueYeon Chung},
booktitle={Conference on Parsimony and Learning (Proceedings Track)},
year={2023},
url={https://openreview.net/forum?id=MxBS6aw5Gd}
}

In recent years, growth in our understanding of the computations performed in both biological and artificial neural networks has largely been driven by either low-level mechanistic studies or global normative approaches. However, concrete methodologies for bridging the gap between these levels of abstraction remain elusive. In this work, we investigate the internal mechanisms of neural networks through the lens of neural population geometry, aiming to provide understanding at an intermediate level of abstraction, as a way to bridge that gap. Utilizing manifold capacity theory (MCT) from statistical physics and manifold alignment analysis (MAA) from high-dimensional statistics, we probe the underlying organization of task-dependent manifolds in deep neural networks and neural recordings from the macaque visual cortex. Specifically, we quantitatively characterize how different learning objectives lead to differences in the organizational strategies of these models and demonstrate how these geometric analyses are connected to the decodability of task-relevant information. Furthermore, these metrics show that macaque visual cortex data are more similar to unsupervised DNNs in terms of geometrical properties such as manifold position and manifold alignment. These analyses present a strong direction for bridging mechanistic and normative theories in neural networks through neural population geometry, potentially opening up many future research avenues in both machine learning and neuroscience.

@phdthesis{chou2023computational,
  title={The Computational Lens: from Quantum Physics to Neuroscience},
  author={Chou, Chi-Ning},
  year={2023},
  school={Harvard University}
}

Two transformative waves of computing have redefined the way we approach science. The first wave came with the birth of the digital computer, which enabled scientists to numerically simulate their models and analyze massive datasets. This technological breakthrough led to the emergence of many sub-disciplines bearing the prefix "computational" in their names. Currently, we are in the midst of the second wave, marked by the remarkable advancements in artificial intelligence. From predicting protein structures to classifying galaxies, the scope of its applications is vast, and there can only be more awaiting us on the horizon.

While these two waves influence scientific methodology at the instrumental level, in this dissertation, I will present the computational lens in science, aiming at the conceptual level. Specifically, the central thesis posits that computation serves as a convenient and mechanistic language for understanding and analyzing information processing systems, offering the advantages of composability and modularity.

This dissertation begins with an illustration of the blueprint of the computational lens, supported by a review of relevant previous work. Subsequently, I will present my own works in quantum physics and neuroscience as concrete examples. In the concluding chapter, I will contemplate the potential of applying the computational lens across various scientific fields, in a way that can provide significant domain insights, and discuss potential future directions.

@journal{paraouty2023sensory,
  title={Sensory cortex plasticity supports auditory social learning},
  author={Paraouty, Nihaad and Yao, Justin D and Varnet, L{\'e}o and Chou, Chi-Ning and Chung, SueYeon and Sanes, Dan H},
  journal={Nature Communications},
  volume={14},
  number={1},
  pages={5828},
  year={2023},
  publisher={Nature Publishing Group UK London}
}

Social learning (SL) through experience with conspecifics can facilitate the acquisition of many behaviors. Thus, when Mongolian gerbils are exposed to a demonstrator performing an auditory discrimination task, their subsequent task acquisition is facilitated, even in the absence of visual cues. Here, we show that transient inactivation of auditory cortex (AC) during exposure caused a significant delay in task acquisition during the subsequent practice phase, suggesting that AC activity is necessary for SL. Moreover, social exposure induced an improvement in AC neuron sensitivity to auditory task cues. The magnitude of neural change during exposure correlated with task acquisition during practice. In contrast, exposure to only auditory task cues led to poorer neurometric and behavioral outcomes. Finally, social information during exposure was encoded in the AC of observer animals. Together, our results suggest that auditory SL is supported by AC neuron plasticity occurring during social exposure and prior to behavioral performance.

@article{LLCCLB22,
  title={A Superconducting Nanowire-based Architecture for Neuromorphic Computing},
  author={Lombo, Andres E. and Lares, Jesus E. and Castellani, Matteo and Chou, Chi-Ning and Lynch, Nancy and Berggren, Karl K.},
  journal={Neuromorphic Computing and Engineering},
  year={2022},
  publisher={IOP Publishing},
  url={http://iopscience.iop.org/article/10.1088/2634-4386/ac86ef}
}

Neuromorphic computing is poised to further the success of software-based neural networks by utilizing improved customized hardware. However, the translation of neuromorphic algorithms to hardware specifications is a problem that has been seldom explored. Building superconducting neuromorphic systems requires extensive expertise in both superconducting physics and theoretical neuroscience. In this work, we aim to bridge this gap by presenting a tool and methodology to translate algorithmic parameters into circuit specifications. We first show the correspondence between theoretical neuroscience models and the dynamics of our circuit topologies. We then apply this tool to solve linear systems by implementing a spiking neural network with our superconducting nanowire-based hardware.

@InProceedings{CLSS22,
  author =	{Chou, Chi-Ning and Love, Peter J. and Sandhu, Juspreet Singh and Shi, Jonathan},
  title =	{Limitations of Local Quantum Algorithms on Random MAX-k-XOR and Beyond},
  booktitle =	{49th International Colloquium on Automata, Languages, and Programming (ICALP 2022)},
  pages =	{41:1--41:20},
  year =	{2022}
}

In this work, we study the limitations of the Quantum Approximate Optimization Algorithm (QAOA) through the lens of statistical physics and show that there exists $\epsilon > 0$, such that $\epsilon \log(n)$ depth QAOA cannot arbitrarily-well approximate the ground state energy of random diluted $k$-spin glasses when $k \geq 4$ is even. This is equivalent to the weak approximation resistance of logarithmic depth QAOA to the Max-$k$-XOR problem. We further extend the limitation to other boolean constraint satisfaction problems as long as the problem satisfies a combinatorial property called the coupled overlap-gap property (OGP) [Chen et al., Annals of Probability, 47(3), 2019]. As a consequence of our techniques, we confirm a conjecture of Brandao et al. [arXiv:1812.04170, 2018] asserting that the landscape independence of QAOA extends to logarithmic depth —-- in other words, for every fixed choice of QAOA angle parameters, the algorithm at logarithmic depth performs almost equally well on almost all instances. Our results provide a new way to study the power and limit of QAOA through statistical physics methods and combinatorial properties.

@article{CGSVV21,
  title={Linear Space Streaming Lower Bounds for Approximating CSPs},
  author={Chou, Chi-Ning and Golovnev, Alexander and Sudan, Madhu and Velingker, Ameya and Velusamy, Santhoshini},
  journal={arXiv preprint},
  year={2021},
  archivePrefix = "arXiv",
  eprint = "2106.13078",
  primaryClass={cs.CC}
}

We consider the approximability of constraint satisfaction problems in the streaming setting. For every constraint satisfaction problem (CSP) on n variables taking values in $\{0,\dots,q−1\}$, we prove that improving over the trivial approximability by a factor of $q$ requires $\Omega(n)$ space even on instances with $O(n)$ constraints. We also identify a broad subclass of problems for which any improvement over the trivial approximability requires $\Omega(n)$ space. The key technical core is an optimal, $q−(k−1)$-inapproximability for the case where every constraint is given by a system of $k−1$ linear equations mod $q$ over $k$ variables. Prior to our work, no such hardness was known for an approximation factor less than $1/2$ for any CSP. Our work builds on and extends the work of Kapralov and Krachun (Proc. STOC 2019) who showed a linear lower bound on any non-trivial approximation of the max cut in graphs. This corresponds roughly to the case of Max $k$-LIN-modq with $k=q=2$. Each one of the extensions provides non-trivial technical challenges that we overcome in this work.

@article{YCC22,
  title={Understanding Rare Spurious Correlations in Neural Networks},
  author={Yang, Yao-Yuan and Chou, Chi-Ning and Chaudhuri, Kamalika},
  eprint={2202.05189},
  archivePrefix={arXiv},
  primaryClass={cs.LG}
}

Neural networks are known to use spurious correlations such as background information for classification. While prior work has looked at spurious correlations that are widespread in the training data, in this work, we investigate how sensitive neural networks are to rare spurious correlations, which may be harder to detect and correct, and may lead to privacy leaks. We introduce spurious patterns correlated with a fixed class to a few training examples and find that it takes only a handful of such examples for the network to learn the correlation. Furthermore, these rare spurious correlations also impact accuracy and privacy. We empirically and theoretically analyze different factors involved in rare spurious correlations and propose mitigation methods accordingly. Specifically, we observe that $\ell_2$ regularization and adding Gaussian noise to inputs can reduce the undesirable effects. Code available at [this https URL](https://github.com/yangarbiter/rare-spurious-correlation){:target="_blank"}.

@article{CGSSV22,
  title={Sketching Approximability of (Weak) Monarchy Predicates},
  author={Chou, Chi-Ning and Golovnev, Alexander and Shahrasbi, Amirbehshad and Sudan, Madhu and Velusamy, Santhoshini},
  eprint={2205.02345},
  archivePrefix={arXiv},
  primaryClass={cs.CC}
}

We analyze the sketching approximability of constraint satisfaction problems on Boolean domains, where the constraints are balanced linear threshold functions applied to literals. In particular, we explore the approximability of monarchy-like functions where the value of the function is determined by a weighted combination of the vote of the first variable (the president) and the sum of the votes of all remaining variables. The pure version of this function is when the president can only be overruled by when all remaining variables agree. For every $k \geq 5$, we show that CSPs where the underlying predicate is a pure monarchy function on $k$ variables have no non-trivial sketching approximation algorithm in $o(\sqrt{n})$ space. We also show infinitely many weaker monarchy functions for which CSPs using such constraints are non-trivially approximable by $O(\log(n))$ space sketching algorithms. Moreover, we give the first example of sketching approximable asymmetric Boolean CSPs.

Our results work within the framework of Chou, Golovnev, Sudan, and Velusamy (FOCS 2021) that characterizes the sketching approximability of all CSPs. Their framework can be applied naturally to get a computer-aided analysis of the approximability of any specific constraint satisfaction problem. The novelty of our work is in using their work to get an analysis that applies to infinitely many problems simultaneously.

@inproceedings{CCZZ21,
  title={Quantum Meets the Minimum Circuit Size Problem},
  author={Chia, Nai-Hui and Chou, Chi-Ning and Zhang, Jiayu and Zhang, Ruizhe},
  booktitle={13th Innovations in Theoretical Computer Science Conference (ITCS 2022)},
  year={2022}
}

In this work, we initiate the study of the Minimum Circuit Size Problem (MCSP) in the quantum setting. MCSP is a problem to compute the circuit complexity of Boolean functions. It is a fascinating problem in complexity theory---its hardness is mysterious, and a better understanding of its hardness can have surprising implications to many fields in computer science.

We first define and investigate the basic complexity-theoretic properties of minimum quantum circuit size problems for three natural objects: Boolean functions, unitaries, and quantum states. We show that these problems are not trivially in NP but in QCMA (or have QCMA protocols). Next, we explore the relations between the three quantum MCSPs and their variants. We discover that some reductions that are not known for classical MCSP exist for quantum MCSPs for unitaries and states, e.g., search-to-decision reduction and self-reduction. Finally, we systematically generalize results known for classical MCSP to the quantum setting (including quantum cryptography, quantum learning theory, quantum circuit lower bounds, and quantum fine-grained complexity) and also find new connections to tomography and quantum gravity. Due to the fundamental differences between classical and quantum circuits, most of our results require extra care and reveal properties and phenomena unique to the quantum setting. Our findings could be of interest for future studies, and we post several open problems for further exploration along this direction.

@article{CGSV21b,
  title={Approximability of all finite CSPs in the dynamic streaming setting},
  author={Chou, Chi-Ning and Golovnev, Alexander and Sudan, Madhu and Velusamy, Santhoshini},
  journal={arXiv preprint},
  year={2021},
  archivePrefix = "arXiv",
  eprint = "2105.01161",
  primaryClass={cs.CC}
}

A constraint satisfaction problem (CSP), Max-CSP($\mathcal{F}$), is specified by a finite set of constraints $\mathcal{F}\subseteq\{[q]^k\to\{0,1\}\}$ for positive integers $q$ and $k$. An instance of the problem on n variables is given by m applications of constraints from $\mathcal{F}$ to subsequences of the n variables, and the goal is to find an assignment to the variables that satisfies the maximum number of constraints. In the $(\gamma,\beta)$-approximation version of the problem for parameters $0\leq\beta\leq\gamma\leq1$, the goal is to distinguish instances where at least $\gamma$ fraction of the constraints can be satisfied from instances where at most $\beta$ fraction of the constraints can be satisfied.

In this work we consider the approximability of this problem in the context of sketching algorithms and give a dichotomy result. Specifically, for every family $\cF$ and every $\beta < \gamma$, we show that either a linear sketching algorithm solves the problem in polylogarithmic space, or the problem is not solvable by any sketching algorithm in $o(\sqrt{n})$ space.

@article{CGSV21,
  title={Approximability of all Boolean CSPs in the dynamic streaming setting},
  author={Chou, Chi-Ning and Golovnev, Sasha and Sudan, Madhu and Velusamy, Santhoshini},
  year={2021},
  archivePrefix = "arXiv",
  eprint = "2102.12351",
  primaryClass={cs.CC}
}

A Boolean constraint satisfaction problem (CSP), $\textsf{Max-CSP}(f)$, is a maximization problem specified by a constraint $f:\{-1,1\}^k\to\{0,1\}$. An instance of the problem consists of $m$ constraint applications on $n$ Boolean variables, where each constraint application applies the constraint to $k$ literals chosen from the $n$ variables and their negations. The goal is to compute the maximum number of constraints that can be satisfied by a Boolean assignment to the $n$~variables. In the $(\gamma,\beta)$-approximation version of the problem for parameters $\gamma \geq \beta \in [0,1]$, the goal is to distinguish instances where at least $\gamma$ fraction of the constraints can be satisfied from instances where at most $\beta$ fraction of the constraints can be satisfied.

In this work we consider the approximability of $\textsf{Max-CSP}(f)$ in the (dynamic) streaming setting, where constraints are inserted (and may also be deleted in the dynamic setting) one at a time. We completely characterize the approximability of all Boolean CSPs in the dynamic streaming setting. Specifically, given $f$, $\gamma$ and $\beta$ we show that either (1) the $(\gamma,\beta)$-approximation version of $\textsf{Max-CSP}(f)$ has a probabilistic dynamic streaming algorithm using $O(\log n)$ space, or (2) for every $\epsilon > 0$ the $(\gamma-\epsilon,\beta+\epsilon)$-approximation version of $\textsf{Max-CSP}(f)$ requires $\Omega(\sqrt{n})$ space for probabilistic dynamic streaming algorithms. We also extend previously known results in the insertion-only setting to a wide variety of cases, and in particular the case of $k=2$ where we get a dichotomy and the case when the satisfying assignments of $f$ support a distribution on $\{-1,1\}^k$ with uniform marginals.

Our positive results show wider applicability of bias-based algorithms used previously by [Guruswami-Velingker-Velusamy, APPROX 2017]] and [Chou-Golovnev-Velusamy, FOCS 2020] by giving a systematic way to discover biases. Our negative results combine the Fourier analytic methods of [Kapralov-Khannan-Sudan, SODA 2015], which we extend to a wider class of CSPs, with a rich collection of reductions among communication complexity problems that lie at the heart of the negative results.

@InProceedings{BCG21,
  author =	{Boaz Barak and Chi-Ning Chou and Xun Gao},
  title =	{Spoofing Linear Cross-Entropy Benchmarking in Shallow Quantum Circuits},
  booktitle =	{12th Innovations in Theoretical Computer Science Conference (ITCS 2021)},
  pages =	{30:1--30:20},
  year =	{2021},
  archivePrefix = "arXiv",
  eprint = "2005.02421",
  primaryClass={quant-ph}
}

The linear cross-entropy benchmark (Linear XEB) has been used as a test for procedures simulating quantum circuits. Given a quantum circuit C with n inputs and outputs and purported simulator whose output is distributed according to a distribution p over $\{0,1\}^n$, the linear XEB fidelity of the simulator is $\mathcal{F}_C(p)=2^n\mathbb{E}_{x\sim p}q_C(x)-1$ where $q_C(x)$ is the probability that x is output from the distribution C|0n⟩. A trivial simulator (e.g., the uniform distribution) satisfies $\mathcal{F}_C(p)=0$, while Google's noisy quantum simulation of a 53 qubit circuit C achieved a fidelity value of (2.24±0.21)×10−3 (Arute et. al., Nature'19).

In this work we give a classical randomized algorithm that for a given circuit C of depth d with Haar random 2-qubit gates achieves in expectation a fidelity value of $\Omega(\frac{n}{L}\cdot15^{-d})$ in running time 𝗉𝗈𝗅𝗒(n,2L). Here L is the size of the *light cone* of C: the maximum number of input bits that each output bit depends on. In particular, we obtain a polynomial-time algorithm that achieves large fidelity of $\omega(1)$ for depth $O(\sqrt{\log n})$ two-dimensional circuits. To our knowledge, this is the first such result for two dimensional circuits of super-constant depth. Our results can be considered as an evidence that fooling the linear XEB test might be easier than achieving a full simulation of the quantum circuit.

@InProceedings{CGV20,
  title={Optimal Streaming Approximations for all Boolean Max-2CSPs and Max-ksat},
  author={Chou, Chi-Ning and Golovnev, Alexander and Velusamy, Santhoshini},
  booktitle={2020 IEEE 61st Annual Symposium on Foundations of Computer Science (FOCS)},
  year={2020},
  pages={330-341},
  doi={10.1109/FOCS46700.2020.00039},
  archivePrefix = "arXiv",
  eprint = "2004.11796",
  primaryClass={cs.CC}
}

We prove tight upper and lower bounds on approximation ratios of all Boolean Max-2CSP problems in the streaming model. Specifically, for every type of Max-2CSP problem, we give an explicit constant $\alpha$, s.t. for any $\epsilon>0$ (i) there is an $(\alpha-\epsilon)$-streaming approximation using space $O(\log n)$; and (ii) any $(\alpha-\epsilon)$-streaming approximation requires space $\Omega(\sqrt{n})$. This generalizes the celebrated work of [Kapralov, Khanna, Sudan SODA 2015; Kapralov, Krachun STOC 2019], who showed that the optimal approximation ratio for Max-CUT was 1/2.

Prior to this work, the problem of determining this ratio was open for all other Max-2CSPs. Our results are quite surprising for some specific Max-2CSPs. For the Max-DCUT problem, there was a gap between an upper bound of 1/2 and a lower bound of 2/5 [Guruswami, Velingker, Velusamy APPROX 2017]. We show that neither of these bounds is tight, and the optimal ratio for Max-DCUT is 4/9. We also establish that the tight approximation for Max-2SAT is $\sqrt{2}/2$, and for Exact Max-2SAT it is 3/4. As a byproduct, our result gives a separation between space-efficient approximations for Max-2SAT and Exact Max-2SAT. This is in sharp contrast to the setting of polynomial-time algorithms with polynomial space, where the two problems are known to be equally hard to approximate.

@InProceedings{CW20,
  title = 	 {ODE-Inspired Analysis for the Biological Version of Oja{'}s Rule in Solving Streaming PCA},
  author = 	 {Chou, Chi-Ning and Wang, Mien Brabeeba},
  booktitle = 	 {Proceedings of Thirty Third Conference on Learning Theory (COLT 2020)},
  pages = 	 {1339--1343},
  year = 	 {2020},
  editor = 	 {Abernethy, Jacob and Agarwal, Shivani},
  volume = 	 {125},
  series = 	 {Proceedings of Machine Learning Research},
  month = 	 {09--12 Jul},
  publisher = 	 {PMLR},
  pdf = 	 {http://proceedings.mlr.press/v125/chou20a/chou20a.pdf},
  url = 	 {http://proceedings.mlr.press/v125/chou20a.html},
  archivePrefix = "arXiv",
  eprint = "1911.02363",
  primaryClass={q-bio.NC}
}

Oja's rule [Oja, Journal of mathematical biology 1982] is a well-known biologically-plausible algorithm using a Hebbian-type synaptic update rule to solve streaming principal component analysis (PCA). Computational neuroscientists have known that this biological version of Oja's rule converges to the top eigenvector of the covariance matrix of the input in the limit. However, prior to this work, it was open to prove any convergence rate guarantee.

In this work, we give the first convergence rate analysis for the biological version of Oja's rule in solving streaming PCA. Moreover, our convergence rate matches the information theoretical lower bound up to logarithmic factors and outperforms the state-of-the-art upper bound for streaming PCA. Furthermore, we develop a novel framework inspired by ordinary differential equations (ODE) to analyze general stochastic dynamics. The framework abandons the traditional step-by-step analysis and instead analyzes a stochastic dynamic in one-shot by giving a closed-form solution to the entire dynamic. The one-shot framework allows us to apply stopping time and martingale techniques to have a flexible and precise control on the dynamic. We believe that this general framework is powerful and should lead to effective yet simple analysis for a large class of problems with stochastic dynamics.

@article{CSWY21,
  title={A General Framework for Analyzing Stochastic Dynamics in Learning Algorithms},
  author={Chou, Chi-Ning and Sandhu, Juspreet Singh and Wang, Mien Brabeeba and Yu, Tiancheng},
  eprint={2006.06171},
  archivePrefix={arXiv},
  primaryClass={math.OC}
}

One of the challenges in analyzing a learning algorithm is the circular entanglement between the objective value and the stochastic noise. This is also known as the "chicken and egg" phenomenon. Traditionally, people tackle this issue with the special structure of the problem and hence the analysis is difficult to generalize.

In this paper, we present a general framework for analyzing high-probability bounds for stochastic dynamics in learning algorithms. Our framework composes standard techniques from probability theory to give a streamlined three-step recipe with a general and flexible principle to tackle the "chicken and egg" problem. We demonstrate the power and the flexibility of our framework by giving unifying analysis for three very different learning problems with both the last iterate and the strong uniform high probability convergence guarantee. The problems are stochastic gradient descent for strongly convex functions, streaming principal component analysis and linear bandit with stochastic gradient descent updates. We either improve or match the state-of-the-art bounds on all three dynamics.

@InProceedings{BCLSS19,
  title = {(Nearly) Efficient Algorithms for the Graph Matching Problem on Correlated Random Graphs},
  author = {Barak, Boaz and Chou, Chi-Ning and Lei, Zhixian and Schramm, Tselil and Sheng, Yueqi},
  booktitle = {Advances in Neural Information Processing Systems 32 (NeurIPS 2019)},
  pages = {9186--9194},
  year = {2019},
  url = {http://papers.nips.cc/paper/9118-nearly-efficient-algorithms-for-the-graph-matching-problem-on-correlated-random-graphs.pdf},
  archivePrefix = "arXiv",
  eprint = "1805.02349",
  primaryClass={cs.DS}
}

We consider the graph matching/similarity problem of determining how similar two given graphs $G_0,G_1$ are and recovering the permutation $\pi$ on the vertices of $G_1$ that minimizes the symmetric difference between the edges of $G_0$ and $\pi(G_1)$. Graph matching/similarity has applications for pattern matching, computer vision, social network anonymization, malware analysis, and more. We give the first efficient algorithms proven to succeed in the correlated Erdös-Rényi model (Pedarsani and Grossglauser, 2011).

Specifically, we give a polynomial time algorithm for the graph similarity/hypothesis testing task which works for every constant level of correlation between the two graphs that can be arbitrarily close to zero. We also give a quasi-polynomial ($n^{O(\log n)}$ time) algorithm for the graph matching task of recovering the permutation minimizing the symmetric difference in this model. This is the first algorithm to do so without requiring as additional input a "seed" of the values of the ground truth permutation on at least $n^{\Omega(1)}$ vertices. Our algorithms follow a general framework of counting the occurrences of subgraphs from a particular family of graphs allowing for tradeoffs between efficiency and accuracy.

@journal{CKS19J,
  author = {Chou, Chi-Ning and Kumar, Mrinal and Solomon, Noam},
  title = {Closure Results for Polynomial Factorization},
  year = {2019},
  pages = {1--34},
  doi = {10.4086/toc.2019.v015a013},
  publisher = {Theory of Computing},
  journal = {Theory of Computing},
  volume = {15},
  number = {13},
  URL = {http://www.theoryofcomputing.org/articles/v015a013},
  archivePrefix = "arXiv",
  eprint = "1803.05933",
  primaryClass={cs.CC}
}

In a sequence of fundamental results in the 1980s, Kaltofen (SICOMP 1985, STOC'86, STOC'87, RANDOM'89) showed that factors of multivariate polynomials with small arithmetic circuits have small arithmetic circuits. In other words, the complexity class $\VP$ is closed under taking factors. A natural question in this context is to understand if other natural classes of multivariate polynomials, for instance, arithmetic formulas, algebraic branching programs, bounded-depth arithmetic circuits or the class $\VNP$, are closed under taking factors.

In this paper, we show that all factors of degree $\log^a n$ of polynomials with $\poly(n)$-size depth-$k$ circuits have $\poly(n)$-size circuits of depth $O(k + a)$. This partially answers a question of Shpilka--Yehudayoff (Found. Trends in TCS, 2010) and has applications to hardness-randomness tradeoffs for bounded-depth arithmetic circuits.

As direct applications of our techniques, we also obtain simple proofs of the following results.

$\bullet$ The complexity class $\VNP$ is closed under taking factors. This confirms Conjecture~2.1 in Bürgisser's monograph (2000) and improves upon a recent result of Dutta, Saxena and Sinhababu (STOC'18) who showed a quasipolynomial upper bound on the number of auxiliary variables and the complexity of the verifier circuit of factors of polynomials in $\VNP$.

$\bullet$ A factor of degree $d$ of a polynomial $P$ which can be computed by an arithmetic formula (or an algebraic branching program) of size $s$ has a formula (an algebraic branching program, resp.) of size $\poly(s, d^{\log d}, \deg(P))$. This result was first shown by Dutta et al. (STOC'18) and we obtain a slightly different proof as an easy consequence of our techniques.

Our proofs rely on a combination of the ideas, based on Hensel lifting, developed in the polynomial factoring literature, and the depth-reduction results for arithmetic circuits, and hold over fields of characteristic zero or of sufficiently large characteristic.

@InProceedings{CLN19,
  author =	{Chi-Ning Chou and Zhixian Lei and Preetum Nakkiran},
  title =	{Tracking the l_2 Norm with Constant Update Time},
  booktitle =	{Approximation, Randomization, and Combinatorial Optimization. Algorithms and Techniques (APPROX/RANDOM 2019)},
  pages =	{2:1--2:15},
  ISSN =	{1868-8969},
  year =	{2019},
  URL =		{http://drops.dagstuhl.de/opus/volltexte/2019/11217},
  URN =		{urn:nbn:de:0030-drops-112175},
  annote =	{Keywords: Streaming algorithms, Sketching algorithms, Tracking, CountSketch},
  archivePrefix = "arXiv",
  eprint = "1807.06479",
  primaryClass={cs.DS}
}

The $\ell_2$ tracking problem is the task of obtaining a streaming algorithm that, given access to a stream of items $a_1,a_2,a_3,\ldots$ from a universe $[n]$, outputs at each time $t$ an estimate to the $\ell_2$ norm of the frequency vector $f^{(t)}\in \mathbb{R}^n$ (where $f^{(t)}_i$ is the number of occurrences of item $i$ in the stream up to time $t$). The previous work [Braverman-Chestnut-Ivkin-Nelson-Wang-Woodruff, PODS 2017] gave an streaming algorithm with (the optimal) space using $O(\epsilon^{-2}\log(1/\delta))$ words and $O(\epsilon^{-2}\log(1/\delta))$ update time to obtain an $\epsilon$-accurate estimate with probability at least $1-\delta$.

We give the first algorithm that achieves update time of $O(\log 1/\delta)$ which is independent of the accuracy parameter $\epsilon$, together with the nearly optimal space using $O(\epsilon^{-2}\log(1/\delta))$ words. Our algorithm is obtained using the Count Sketch of [Charilkar-Chen-Farach-Colton, ICALP 2002].

@unpublished{CKS19,
  title={Closure of VP under taking factors: a short and simple proof},
  author={Chou, Chi-Ning and Kumar, Mrinal and Solomon, Noam},
  note={In submission to Theory of Computing},
  eprint={1903.02366},
  archivePrefix={arXiv},
  primaryClass={cs.CC}
}

In this note, we give a short, simple and almost completely self contained proof of a classical result of Kaltofen which shows that if an $n$ variate degree $d$ polynomial $f$ can be computed by an arithmetic circuit of size $s$, then each of its factors can be computed by an arithmetic circuit of size at most $\poly\left(s, n, d\right)$.

However, unlike Kaltofen's argument, our proof does not directly give an efficient algorithm for computing the circuits for the factors of $f$.

@InProceedings{CCL19,
  author ={Chi-Ning Chou and Kai-Min Chung and Chi-Jen Lu},
  title ={On the Algorithmic Power of Spiking Neural Networks},
  booktitle ={10th Innovations in Theoretical Computer Science  Conference (ITCS 2019)},
  pages ={26:1--26:20},
  year ={2019},
  URL ={http://drops.dagstuhl.de/opus/volltexte/2018/10119},
  archivePrefix = "arXiv",
  eprint = "1803.10375",
  primaryClass={cs.NE}
}

@InProceedings{CLCCTL18,
  title={Personalized Difficulty Adjustment for Countering the Double-Spending Attack in Proof-of-Work Consensus Protocols},
  author={Chou, Chi-Ning and Lin, Yu-Jing and Chen, Ren and Chang, Hsiu-Yao and Tu, I-Ping and Liao, Shih-wei},
  booktitle={2018 IEEE Conference on Blockchain (Blockchain 2018)},
  pages={1457-1462},
  year={2018}
}

@InProceedings{CKS18,
  author ={Chi-Ning Chou and Mrinal Kumar and Noam Solomon},
  title ={Hardness vs Randomness for Bounded Depth Arithmetic Circuits},
  booktitle ={33rd Computational Complexity Conference (CCC 2018)},
  pages ={13:1--13:17},
  ISSN ={1868-8969},
  year ={2018},
  URL ={http://drops.dagstuhl.de/opus/volltexte/2018/8876},
  URN ={urn:nbn:de:0030-drops-88765},
  annote ={Keywords: Algebraic Complexity, Polynomial Factorization Circuit Lower Bounds, Polynomial Identity Testing}
}

In this paper, we study the question of hardness-randomness tradeoffs for bounded depth arithmetic circuits. We show that if there is a family of explicit polynomials $\{f_n\}$, where $f_n$ is of degree $O(\log^2n/\log^2\log n)$ in $n$ variables such that $f_n$ cannot be computed by a depth $\Delta$ arithmetic circuits of size $\poly(n)$, then there is a deterministic sub-exponential time algorithm for polynomial identity testing of arithmetic circuits of depth $\Delta-5$.

This is incomparable to a beautiful result of Dvir et al.[SICOMP, 2009], where they showed that super-polynomial lower bounds for depth $\Delta$ circuits for any explicit family of polynomials (of potentially high degree) implies sub-exponential time deterministic PIT for depth $\Delta-5$ circuits of bounded individual degree. Thus, we remove the ``bounded individual degree" condition in the work of Dvir et al. at the cost of strengthening the hardness assumption to hold for polynomials of low degree.

The key technical ingredient of our proof is the following property of roots of polynomials computable by a bounded depth arithmetic circuit : if $f(x_1, x_2, \ldots, x_n)$ and $P(x_1, x_2, \ldots, x_n, y)$ are polynomials of degree $d$ and $r$ respectively, such that $P$ can be computed by a circuit of size $s$ and depth $\Delta$ and $P(x_1, x_2, \ldots, x_n, f) \equiv 0$, then, $f$ can be computed by a circuit of size $\poly(n, s, r, d^{O(\sqrt{d})})$ and depth $\Delta + 3$. In comparison, Dvir et al. showed that $f$ can be computed by a circuit of depth $\Delta + 3$ and size $\poly(n, s, r, d^{t})$, where $t$ is the degree of $P$ in $y$. Thus, the size upper bound in the work of Dvir et al. is non-trivial when $t$ is small but $d$ could be large, whereas our size upper bound is non-trivial when $d$ is small, but $t$ could be large.

@phdthesis{chou2023computational,
  title={The Computational Lens: from Quantum Physics to Neuroscience},
  author={Chou, Chi-Ning},
  year={2023},
  school={Harvard University}
}

Two transformative waves of computing have redefined the way we approach science. The first wave came with the birth of the digital computer, which enabled scientists to numerically simulate their models and analyze massive datasets. This technological breakthrough led to the emergence of many sub-disciplines bearing the prefix "computational" in their names. Currently, we are in the midst of the second wave, marked by the remarkable advancements in artificial intelligence. From predicting protein structures to classifying galaxies, the scope of its applications is vast, and there can only be more awaiting us on the horizon.

While these two waves influence scientific methodology at the instrumental level, in this dissertation, I will present the computational lens in science, aiming at the conceptual level. Specifically, the central thesis posits that computation serves as a convenient and mechanistic language for understanding and analyzing information processing systems, offering the advantages of composability and modularity.

This dissertation begins with an illustration of the blueprint of the computational lens, supported by a review of relevant previous work. Subsequently, I will present my own works in quantum physics and neuroscience as concrete examples. In the concluding chapter, I will contemplate the potential of applying the computational lens across various scientific fields, in a way that can provide significant domain insights, and discuss potential future directions.

@article{PRXQuantum.5.010334,
  title = {Limitations of Linear Cross-Entropy as a Measure for Quantum Advantage},
  author = {Gao, Xun and Kalinowski, Marcin and Chou, Chi-Ning and Lukin, Mikhail D. and Barak, Boaz and Choi, Soonwon},
  journal = {PRX Quantum},
  volume = {5},
  issue = {1},
  pages = {010334},
  numpages = {27},
  year = {2024},
  month = {Feb},
  publisher = {American Physical Society},
  doi = {10.1103/PRXQuantum.5.010334},
  url = {https://link.aps.org/doi/10.1103/PRXQuantum.5.010334}
}

Demonstrating quantum advantage requires experimental implementation of a computational task that is hard to achieve using state-of-the-art classical systems. One approach is to perform sampling from a probability distribution associated with a class of highly entangled many-body wavefunctions. It has been suggested that this approach can be certified with the Linear Cross-Entropy Benchmark (XEB). We critically examine this notion. First, in a "benign" setting where an honest implementation of noisy quantum circuits is assumed, we characterize the conditions under which the XEB approximates the fidelity. Second, in an "adversarial" setting where all possible classical algorithms are considered for comparison, we show that achieving relatively high XEB values does not imply faithful simulation of quantum dynamics. We present an efficient classical algorithm that, with 1 GPU within 2s, yields high XEB values, namely 2-12% of those obtained in experiments. By identifying and exploiting several vulnerabilities of the XEB, we achieve high XEB values without full simulation of quantum circuits. Remarkably, our algorithm features better scaling with the system size than noisy quantum devices for commonly studied random circuit ensembles. To quantitatively explain the success of our algorithm and the limitations of the XEB, we use a theoretical framework in which the average XEB and fidelity are mapped to statistical models. We illustrate the relation between the XEB and the fidelity for quantum circuits in various architectures, with different gate choices, and in the presence of noise. Our results show that XEB's utility as a proxy for fidelity hinges on several conditions, which must be checked in the benign setting but cannot be assumed in the adversarial setting. Thus, the XEB alone has limited utility as a benchmark for quantum advantage. We discuss ways to overcome these limitations.

@journal{CGSV24J,
  author = {Chou, Chi-Ning and Golovnev, Alexander and Sudan, Madhu and Velusamy, Santhoshini},
  title = {Sketching Approximability of All Finite CSPs},
  year = {2024},
  publisher = {Association for Computing Machinery},
  address = {New York, NY, USA},
  issn = {0004-5411},
  url = {https://doi.org/10.1145/3649435},
  doi = {10.1145/3649435},
  journal = {J. ACM},
}

A constraint satisfaction problem (CSP), $\textsf{Max-CSP}(\mathcal{F})$, is specified by a finite set of constraints $\mathcal{F} \subseteq \{[q]^k \to \{0,1\}\}$ for positive integers $q$ and $k$. An instance of the problem on $n$ variables is given by $m$ applications of constraints from $\mathcal{F}$ to subsequences of the $n$ variables, and the goal is to find an assignment to the variables that satisfies the maximum number of constraints. In the $(\gamma,\beta)$-approximation version of the problem for parameters $0 \leq \beta < \gamma \leq 1$, the goal is to distinguish instances where at least $\gamma$ fraction of the constraints can be satisfied from instances where at most $\beta$ fraction of the constraints can be satisfied. In this work we consider the approximability of this problem in the context of sketching algorithms and give a dichotomy result. Specifically, for every family $\mathcal{F}$ and every $\beta < \gamma$, we show that either a linear sketching algorithm solves the problem in polylogarithmic space, or the problem is not solvable by any sketching algorithm in $o(\sqrt{n})$ space.

@journal{paraouty2023sensory,
  title={Sensory cortex plasticity supports auditory social learning},
  author={Paraouty, Nihaad and Yao, Justin D and Varnet, L{\'e}o and Chou, Chi-Ning and Chung, SueYeon and Sanes, Dan H},
  journal={Nature Communications},
  volume={14},
  number={1},
  pages={5828},
  year={2023},
  publisher={Nature Publishing Group UK London}
}

Social learning (SL) through experience with conspecifics can facilitate the acquisition of many behaviors. Thus, when Mongolian gerbils are exposed to a demonstrator performing an auditory discrimination task, their subsequent task acquisition is facilitated, even in the absence of visual cues. Here, we show that transient inactivation of auditory cortex (AC) during exposure caused a significant delay in task acquisition during the subsequent practice phase, suggesting that AC activity is necessary for SL. Moreover, social exposure induced an improvement in AC neuron sensitivity to auditory task cues. The magnitude of neural change during exposure correlated with task acquisition during practice. In contrast, exposure to only auditory task cues led to poorer neurometric and behavioral outcomes. Finally, social information during exposure was encoded in the AC of observer animals. Together, our results suggest that auditory SL is supported by AC neuron plasticity occurring during social exposure and prior to behavioral performance.

@article{LLCCLB22,
  title={A Superconducting Nanowire-based Architecture for Neuromorphic Computing},
  author={Lombo, Andres E. and Lares, Jesus E. and Castellani, Matteo and Chou, Chi-Ning and Lynch, Nancy and Berggren, Karl K.},
  journal={Neuromorphic Computing and Engineering},
  year={2022},
  publisher={IOP Publishing},
  url={http://iopscience.iop.org/article/10.1088/2634-4386/ac86ef}
}

Neuromorphic computing is poised to further the success of software-based neural networks by utilizing improved customized hardware. However, the translation of neuromorphic algorithms to hardware specifications is a problem that has been seldom explored. Building superconducting neuromorphic systems requires extensive expertise in both superconducting physics and theoretical neuroscience. In this work, we aim to bridge this gap by presenting a tool and methodology to translate algorithmic parameters into circuit specifications. We first show the correspondence between theoretical neuroscience models and the dynamics of our circuit topologies. We then apply this tool to solve linear systems by implementing a spiking neural network with our superconducting nanowire-based hardware.

@journal{CKS19J,
  author = {Chou, Chi-Ning and Kumar, Mrinal and Solomon, Noam},
  title = {Closure Results for Polynomial Factorization},
  year = {2019},
  pages = {1--34},
  doi = {10.4086/toc.2019.v015a013},
  publisher = {Theory of Computing},
  journal = {Theory of Computing},
  volume = {15},
  number = {13},
  URL = {http://www.theoryofcomputing.org/articles/v015a013},
  archivePrefix = "arXiv",
  eprint = "1803.05933",
  primaryClass={cs.CC}
}

In a sequence of fundamental results in the 1980s, Kaltofen (SICOMP 1985, STOC'86, STOC'87, RANDOM'89) showed that factors of multivariate polynomials with small arithmetic circuits have small arithmetic circuits. In other words, the complexity class $\VP$ is closed under taking factors. A natural question in this context is to understand if other natural classes of multivariate polynomials, for instance, arithmetic formulas, algebraic branching programs, bounded-depth arithmetic circuits or the class $\VNP$, are closed under taking factors.

In this paper, we show that all factors of degree $\log^a n$ of polynomials with $\poly(n)$-size depth-$k$ circuits have $\poly(n)$-size circuits of depth $O(k + a)$. This partially answers a question of Shpilka--Yehudayoff (Found. Trends in TCS, 2010) and has applications to hardness-randomness tradeoffs for bounded-depth arithmetic circuits.

As direct applications of our techniques, we also obtain simple proofs of the following results.

$\bullet$ The complexity class $\VNP$ is closed under taking factors. This confirms Conjecture~2.1 in Bürgisser's monograph (2000) and improves upon a recent result of Dutta, Saxena and Sinhababu (STOC'18) who showed a quasipolynomial upper bound on the number of auxiliary variables and the complexity of the verifier circuit of factors of polynomials in $\VNP$.

$\bullet$ A factor of degree $d$ of a polynomial $P$ which can be computed by an arithmetic formula (or an algebraic branching program) of size $s$ has a formula (an algebraic branching program, resp.) of size $\poly(s, d^{\log d}, \deg(P))$. This result was first shown by Dutta et al. (STOC'18) and we obtain a slightly different proof as an easy consequence of our techniques.

Our proofs rely on a combination of the ideas, based on Hensel lifting, developed in the polynomial factoring literature, and the depth-reduction results for arithmetic circuits, and hold over fields of characteristic zero or of sufficiently large characteristic.

@inproceedings{KCPDDSC24,
title={Probing Biological and Artificial Neural Networks with Task-dependent Neural Manifolds},
author={Michael Kuoch and Chi-Ning Chou and Nikhil Parthasarathy and Joel Dapello and James J. DiCarlo and Haim Sompolinsky and SueYeon Chung},
booktitle={Conference on Parsimony and Learning (Proceedings Track)},
year={2023},
url={https://openreview.net/forum?id=MxBS6aw5Gd}
}

In recent years, growth in our understanding of the computations performed in both biological and artificial neural networks has largely been driven by either low-level mechanistic studies or global normative approaches. However, concrete methodologies for bridging the gap between these levels of abstraction remain elusive. In this work, we investigate the internal mechanisms of neural networks through the lens of neural population geometry, aiming to provide understanding at an intermediate level of abstraction, as a way to bridge that gap. Utilizing manifold capacity theory (MCT) from statistical physics and manifold alignment analysis (MAA) from high-dimensional statistics, we probe the underlying organization of task-dependent manifolds in deep neural networks and neural recordings from the macaque visual cortex. Specifically, we quantitatively characterize how different learning objectives lead to differences in the organizational strategies of these models and demonstrate how these geometric analyses are connected to the decodability of task-relevant information. Furthermore, these metrics show that macaque visual cortex data are more similar to unsupervised DNNs in terms of geometrical properties such as manifold position and manifold alignment. These analyses present a strong direction for bridging mechanistic and normative theories in neural networks through neural population geometry, potentially opening up many future research avenues in both machine learning and neuroscience.

@InProceedings{CLSS22,
  author =	{Chou, Chi-Ning and Love, Peter J. and Sandhu, Juspreet Singh and Shi, Jonathan},
  title =	{Limitations of Local Quantum Algorithms on Random MAX-k-XOR and Beyond},
  booktitle =	{49th International Colloquium on Automata, Languages, and Programming (ICALP 2022)},
  pages =	{41:1--41:20},
  year =	{2022}
}

In this work, we study the limitations of the Quantum Approximate Optimization Algorithm (QAOA) through the lens of statistical physics and show that there exists $\epsilon > 0$, such that $\epsilon \log(n)$ depth QAOA cannot arbitrarily-well approximate the ground state energy of random diluted $k$-spin glasses when $k \geq 4$ is even. This is equivalent to the weak approximation resistance of logarithmic depth QAOA to the Max-$k$-XOR problem. We further extend the limitation to other boolean constraint satisfaction problems as long as the problem satisfies a combinatorial property called the coupled overlap-gap property (OGP) [Chen et al., Annals of Probability, 47(3), 2019]. As a consequence of our techniques, we confirm a conjecture of Brandao et al. [arXiv:1812.04170, 2018] asserting that the landscape independence of QAOA extends to logarithmic depth —-- in other words, for every fixed choice of QAOA angle parameters, the algorithm at logarithmic depth performs almost equally well on almost all instances. Our results provide a new way to study the power and limit of QAOA through statistical physics methods and combinatorial properties.

@article{CGSVV21,
  title={Linear Space Streaming Lower Bounds for Approximating CSPs},
  author={Chou, Chi-Ning and Golovnev, Alexander and Sudan, Madhu and Velingker, Ameya and Velusamy, Santhoshini},
  journal={arXiv preprint},
  year={2021},
  archivePrefix = "arXiv",
  eprint = "2106.13078",
  primaryClass={cs.CC}
}

We consider the approximability of constraint satisfaction problems in the streaming setting. For every constraint satisfaction problem (CSP) on n variables taking values in $\{0,\dots,q−1\}$, we prove that improving over the trivial approximability by a factor of $q$ requires $\Omega(n)$ space even on instances with $O(n)$ constraints. We also identify a broad subclass of problems for which any improvement over the trivial approximability requires $\Omega(n)$ space. The key technical core is an optimal, $q−(k−1)$-inapproximability for the case where every constraint is given by a system of $k−1$ linear equations mod $q$ over $k$ variables. Prior to our work, no such hardness was known for an approximation factor less than $1/2$ for any CSP. Our work builds on and extends the work of Kapralov and Krachun (Proc. STOC 2019) who showed a linear lower bound on any non-trivial approximation of the max cut in graphs. This corresponds roughly to the case of Max $k$-LIN-modq with $k=q=2$. Each one of the extensions provides non-trivial technical challenges that we overcome in this work.

@article{CGSSV22,
  title={Sketching Approximability of (Weak) Monarchy Predicates},
  author={Chou, Chi-Ning and Golovnev, Alexander and Shahrasbi, Amirbehshad and Sudan, Madhu and Velusamy, Santhoshini},
  eprint={2205.02345},
  archivePrefix={arXiv},
  primaryClass={cs.CC}
}

We analyze the sketching approximability of constraint satisfaction problems on Boolean domains, where the constraints are balanced linear threshold functions applied to literals. In particular, we explore the approximability of monarchy-like functions where the value of the function is determined by a weighted combination of the vote of the first variable (the president) and the sum of the votes of all remaining variables. The pure version of this function is when the president can only be overruled by when all remaining variables agree. For every $k \geq 5$, we show that CSPs where the underlying predicate is a pure monarchy function on $k$ variables have no non-trivial sketching approximation algorithm in $o(\sqrt{n})$ space. We also show infinitely many weaker monarchy functions for which CSPs using such constraints are non-trivially approximable by $O(\log(n))$ space sketching algorithms. Moreover, we give the first example of sketching approximable asymmetric Boolean CSPs.

Our results work within the framework of Chou, Golovnev, Sudan, and Velusamy (FOCS 2021) that characterizes the sketching approximability of all CSPs. Their framework can be applied naturally to get a computer-aided analysis of the approximability of any specific constraint satisfaction problem. The novelty of our work is in using their work to get an analysis that applies to infinitely many problems simultaneously.

@inproceedings{CCZZ21,
  title={Quantum Meets the Minimum Circuit Size Problem},
  author={Chia, Nai-Hui and Chou, Chi-Ning and Zhang, Jiayu and Zhang, Ruizhe},
  booktitle={13th Innovations in Theoretical Computer Science Conference (ITCS 2022)},
  year={2022}
}

In this work, we initiate the study of the Minimum Circuit Size Problem (MCSP) in the quantum setting. MCSP is a problem to compute the circuit complexity of Boolean functions. It is a fascinating problem in complexity theory---its hardness is mysterious, and a better understanding of its hardness can have surprising implications to many fields in computer science.

We first define and investigate the basic complexity-theoretic properties of minimum quantum circuit size problems for three natural objects: Boolean functions, unitaries, and quantum states. We show that these problems are not trivially in NP but in QCMA (or have QCMA protocols). Next, we explore the relations between the three quantum MCSPs and their variants. We discover that some reductions that are not known for classical MCSP exist for quantum MCSPs for unitaries and states, e.g., search-to-decision reduction and self-reduction. Finally, we systematically generalize results known for classical MCSP to the quantum setting (including quantum cryptography, quantum learning theory, quantum circuit lower bounds, and quantum fine-grained complexity) and also find new connections to tomography and quantum gravity. Due to the fundamental differences between classical and quantum circuits, most of our results require extra care and reveal properties and phenomena unique to the quantum setting. Our findings could be of interest for future studies, and we post several open problems for further exploration along this direction.

@article{CGSV21b,
  title={Approximability of all finite CSPs in the dynamic streaming setting},
  author={Chou, Chi-Ning and Golovnev, Alexander and Sudan, Madhu and Velusamy, Santhoshini},
  journal={arXiv preprint},
  year={2021},
  archivePrefix = "arXiv",
  eprint = "2105.01161",
  primaryClass={cs.CC}
}

A constraint satisfaction problem (CSP), Max-CSP($\mathcal{F}$), is specified by a finite set of constraints $\mathcal{F}\subseteq\{[q]^k\to\{0,1\}\}$ for positive integers $q$ and $k$. An instance of the problem on n variables is given by m applications of constraints from $\mathcal{F}$ to subsequences of the n variables, and the goal is to find an assignment to the variables that satisfies the maximum number of constraints. In the $(\gamma,\beta)$-approximation version of the problem for parameters $0\leq\beta\leq\gamma\leq1$, the goal is to distinguish instances where at least $\gamma$ fraction of the constraints can be satisfied from instances where at most $\beta$ fraction of the constraints can be satisfied.

In this work we consider the approximability of this problem in the context of sketching algorithms and give a dichotomy result. Specifically, for every family $\cF$ and every $\beta < \gamma$, we show that either a linear sketching algorithm solves the problem in polylogarithmic space, or the problem is not solvable by any sketching algorithm in $o(\sqrt{n})$ space.

@InProceedings{BCG21,
  author =	{Boaz Barak and Chi-Ning Chou and Xun Gao},
  title =	{Spoofing Linear Cross-Entropy Benchmarking in Shallow Quantum Circuits},
  booktitle =	{12th Innovations in Theoretical Computer Science Conference (ITCS 2021)},
  pages =	{30:1--30:20},
  year =	{2021},
  archivePrefix = "arXiv",
  eprint = "2005.02421",
  primaryClass={quant-ph}
}

The linear cross-entropy benchmark (Linear XEB) has been used as a test for procedures simulating quantum circuits. Given a quantum circuit C with n inputs and outputs and purported simulator whose output is distributed according to a distribution p over $\{0,1\}^n$, the linear XEB fidelity of the simulator is $\mathcal{F}_C(p)=2^n\mathbb{E}_{x\sim p}q_C(x)-1$ where $q_C(x)$ is the probability that x is output from the distribution C|0n⟩. A trivial simulator (e.g., the uniform distribution) satisfies $\mathcal{F}_C(p)=0$, while Google's noisy quantum simulation of a 53 qubit circuit C achieved a fidelity value of (2.24±0.21)×10−3 (Arute et. al., Nature'19).

In this work we give a classical randomized algorithm that for a given circuit C of depth d with Haar random 2-qubit gates achieves in expectation a fidelity value of $\Omega(\frac{n}{L}\cdot15^{-d})$ in running time 𝗉𝗈𝗅𝗒(n,2L). Here L is the size of the *light cone* of C: the maximum number of input bits that each output bit depends on. In particular, we obtain a polynomial-time algorithm that achieves large fidelity of $\omega(1)$ for depth $O(\sqrt{\log n})$ two-dimensional circuits. To our knowledge, this is the first such result for two dimensional circuits of super-constant depth. Our results can be considered as an evidence that fooling the linear XEB test might be easier than achieving a full simulation of the quantum circuit.

@InProceedings{CGV20,
  title={Optimal Streaming Approximations for all Boolean Max-2CSPs and Max-ksat},
  author={Chou, Chi-Ning and Golovnev, Alexander and Velusamy, Santhoshini},
  booktitle={2020 IEEE 61st Annual Symposium on Foundations of Computer Science (FOCS)},
  year={2020},
  pages={330-341},
  doi={10.1109/FOCS46700.2020.00039},
  archivePrefix = "arXiv",
  eprint = "2004.11796",
  primaryClass={cs.CC}
}

We prove tight upper and lower bounds on approximation ratios of all Boolean Max-2CSP problems in the streaming model. Specifically, for every type of Max-2CSP problem, we give an explicit constant $\alpha$, s.t. for any $\epsilon>0$ (i) there is an $(\alpha-\epsilon)$-streaming approximation using space $O(\log n)$; and (ii) any $(\alpha-\epsilon)$-streaming approximation requires space $\Omega(\sqrt{n})$. This generalizes the celebrated work of [Kapralov, Khanna, Sudan SODA 2015; Kapralov, Krachun STOC 2019], who showed that the optimal approximation ratio for Max-CUT was 1/2.

Prior to this work, the problem of determining this ratio was open for all other Max-2CSPs. Our results are quite surprising for some specific Max-2CSPs. For the Max-DCUT problem, there was a gap between an upper bound of 1/2 and a lower bound of 2/5 [Guruswami, Velingker, Velusamy APPROX 2017]. We show that neither of these bounds is tight, and the optimal ratio for Max-DCUT is 4/9. We also establish that the tight approximation for Max-2SAT is $\sqrt{2}/2$, and for Exact Max-2SAT it is 3/4. As a byproduct, our result gives a separation between space-efficient approximations for Max-2SAT and Exact Max-2SAT. This is in sharp contrast to the setting of polynomial-time algorithms with polynomial space, where the two problems are known to be equally hard to approximate.

@InProceedings{CW20,
  title = 	 {ODE-Inspired Analysis for the Biological Version of Oja{'}s Rule in Solving Streaming PCA},
  author = 	 {Chou, Chi-Ning and Wang, Mien Brabeeba},
  booktitle = 	 {Proceedings of Thirty Third Conference on Learning Theory (COLT 2020)},
  pages = 	 {1339--1343},
  year = 	 {2020},
  editor = 	 {Abernethy, Jacob and Agarwal, Shivani},
  volume = 	 {125},
  series = 	 {Proceedings of Machine Learning Research},
  month = 	 {09--12 Jul},
  publisher = 	 {PMLR},
  pdf = 	 {http://proceedings.mlr.press/v125/chou20a/chou20a.pdf},
  url = 	 {http://proceedings.mlr.press/v125/chou20a.html},
  archivePrefix = "arXiv",
  eprint = "1911.02363",
  primaryClass={q-bio.NC}
}

Oja's rule [Oja, Journal of mathematical biology 1982] is a well-known biologically-plausible algorithm using a Hebbian-type synaptic update rule to solve streaming principal component analysis (PCA). Computational neuroscientists have known that this biological version of Oja's rule converges to the top eigenvector of the covariance matrix of the input in the limit. However, prior to this work, it was open to prove any convergence rate guarantee.

In this work, we give the first convergence rate analysis for the biological version of Oja's rule in solving streaming PCA. Moreover, our convergence rate matches the information theoretical lower bound up to logarithmic factors and outperforms the state-of-the-art upper bound for streaming PCA. Furthermore, we develop a novel framework inspired by ordinary differential equations (ODE) to analyze general stochastic dynamics. The framework abandons the traditional step-by-step analysis and instead analyzes a stochastic dynamic in one-shot by giving a closed-form solution to the entire dynamic. The one-shot framework allows us to apply stopping time and martingale techniques to have a flexible and precise control on the dynamic. We believe that this general framework is powerful and should lead to effective yet simple analysis for a large class of problems with stochastic dynamics.

@InProceedings{BCLSS19,
  title = {(Nearly) Efficient Algorithms for the Graph Matching Problem on Correlated Random Graphs},
  author = {Barak, Boaz and Chou, Chi-Ning and Lei, Zhixian and Schramm, Tselil and Sheng, Yueqi},
  booktitle = {Advances in Neural Information Processing Systems 32 (NeurIPS 2019)},
  pages = {9186--9194},
  year = {2019},
  url = {http://papers.nips.cc/paper/9118-nearly-efficient-algorithms-for-the-graph-matching-problem-on-correlated-random-graphs.pdf},
  archivePrefix = "arXiv",
  eprint = "1805.02349",
  primaryClass={cs.DS}
}

We consider the graph matching/similarity problem of determining how similar two given graphs $G_0,G_1$ are and recovering the permutation $\pi$ on the vertices of $G_1$ that minimizes the symmetric difference between the edges of $G_0$ and $\pi(G_1)$. Graph matching/similarity has applications for pattern matching, computer vision, social network anonymization, malware analysis, and more. We give the first efficient algorithms proven to succeed in the correlated Erdös-Rényi model (Pedarsani and Grossglauser, 2011).

Specifically, we give a polynomial time algorithm for the graph similarity/hypothesis testing task which works for every constant level of correlation between the two graphs that can be arbitrarily close to zero. We also give a quasi-polynomial ($n^{O(\log n)}$ time) algorithm for the graph matching task of recovering the permutation minimizing the symmetric difference in this model. This is the first algorithm to do so without requiring as additional input a "seed" of the values of the ground truth permutation on at least $n^{\Omega(1)}$ vertices. Our algorithms follow a general framework of counting the occurrences of subgraphs from a particular family of graphs allowing for tradeoffs between efficiency and accuracy.

@InProceedings{CLN19,
  author =	{Chi-Ning Chou and Zhixian Lei and Preetum Nakkiran},
  title =	{Tracking the l_2 Norm with Constant Update Time},
  booktitle =	{Approximation, Randomization, and Combinatorial Optimization. Algorithms and Techniques (APPROX/RANDOM 2019)},
  pages =	{2:1--2:15},
  ISSN =	{1868-8969},
  year =	{2019},
  URL =		{http://drops.dagstuhl.de/opus/volltexte/2019/11217},
  URN =		{urn:nbn:de:0030-drops-112175},
  annote =	{Keywords: Streaming algorithms, Sketching algorithms, Tracking, CountSketch},
  archivePrefix = "arXiv",
  eprint = "1807.06479",
  primaryClass={cs.DS}
}

The $\ell_2$ tracking problem is the task of obtaining a streaming algorithm that, given access to a stream of items $a_1,a_2,a_3,\ldots$ from a universe $[n]$, outputs at each time $t$ an estimate to the $\ell_2$ norm of the frequency vector $f^{(t)}\in \mathbb{R}^n$ (where $f^{(t)}_i$ is the number of occurrences of item $i$ in the stream up to time $t$). The previous work [Braverman-Chestnut-Ivkin-Nelson-Wang-Woodruff, PODS 2017] gave an streaming algorithm with (the optimal) space using $O(\epsilon^{-2}\log(1/\delta))$ words and $O(\epsilon^{-2}\log(1/\delta))$ update time to obtain an $\epsilon$-accurate estimate with probability at least $1-\delta$.

We give the first algorithm that achieves update time of $O(\log 1/\delta)$ which is independent of the accuracy parameter $\epsilon$, together with the nearly optimal space using $O(\epsilon^{-2}\log(1/\delta))$ words. Our algorithm is obtained using the Count Sketch of [Charilkar-Chen-Farach-Colton, ICALP 2002].

@InProceedings{CCL19,
  author ={Chi-Ning Chou and Kai-Min Chung and Chi-Jen Lu},
  title ={On the Algorithmic Power of Spiking Neural Networks},
  booktitle ={10th Innovations in Theoretical Computer Science  Conference (ITCS 2019)},
  pages ={26:1--26:20},
  year ={2019},
  URL ={http://drops.dagstuhl.de/opus/volltexte/2018/10119},
  archivePrefix = "arXiv",
  eprint = "1803.10375",
  primaryClass={cs.NE}
}

@InProceedings{CLCCTL18,
  title={Personalized Difficulty Adjustment for Countering the Double-Spending Attack in Proof-of-Work Consensus Protocols},
  author={Chou, Chi-Ning and Lin, Yu-Jing and Chen, Ren and Chang, Hsiu-Yao and Tu, I-Ping and Liao, Shih-wei},
  booktitle={2018 IEEE Conference on Blockchain (Blockchain 2018)},
  pages={1457-1462},
  year={2018}
}

@InProceedings{CKS18,
  author ={Chi-Ning Chou and Mrinal Kumar and Noam Solomon},
  title ={Hardness vs Randomness for Bounded Depth Arithmetic Circuits},
  booktitle ={33rd Computational Complexity Conference (CCC 2018)},
  pages ={13:1--13:17},
  ISSN ={1868-8969},
  year ={2018},
  URL ={http://drops.dagstuhl.de/opus/volltexte/2018/8876},
  URN ={urn:nbn:de:0030-drops-88765},
  annote ={Keywords: Algebraic Complexity, Polynomial Factorization Circuit Lower Bounds, Polynomial Identity Testing}
}

In this paper, we study the question of hardness-randomness tradeoffs for bounded depth arithmetic circuits. We show that if there is a family of explicit polynomials $\{f_n\}$, where $f_n$ is of degree $O(\log^2n/\log^2\log n)$ in $n$ variables such that $f_n$ cannot be computed by a depth $\Delta$ arithmetic circuits of size $\poly(n)$, then there is a deterministic sub-exponential time algorithm for polynomial identity testing of arithmetic circuits of depth $\Delta-5$.

This is incomparable to a beautiful result of Dvir et al.[SICOMP, 2009], where they showed that super-polynomial lower bounds for depth $\Delta$ circuits for any explicit family of polynomials (of potentially high degree) implies sub-exponential time deterministic PIT for depth $\Delta-5$ circuits of bounded individual degree. Thus, we remove the ``bounded individual degree" condition in the work of Dvir et al. at the cost of strengthening the hardness assumption to hold for polynomials of low degree.

The key technical ingredient of our proof is the following property of roots of polynomials computable by a bounded depth arithmetic circuit : if $f(x_1, x_2, \ldots, x_n)$ and $P(x_1, x_2, \ldots, x_n, y)$ are polynomials of degree $d$ and $r$ respectively, such that $P$ can be computed by a circuit of size $s$ and depth $\Delta$ and $P(x_1, x_2, \ldots, x_n, f) \equiv 0$, then, $f$ can be computed by a circuit of size $\poly(n, s, r, d^{O(\sqrt{d})})$ and depth $\Delta + 3$. In comparison, Dvir et al. showed that $f$ can be computed by a circuit of depth $\Delta + 3$ and size $\poly(n, s, r, d^{t})$, where $t$ is the degree of $P$ in $y$. Thus, the size upper bound in the work of Dvir et al. is non-trivial when $t$ is small but $d$ could be large, whereas our size upper bound is non-trivial when $d$ is small, but $t$ could be large.

@article {Hu2024.11.17.623906,
  author = {Hu, Bo and Temiz, Nesibe Z. and Chou, Chi-Ning and Rupprecht, Peter and Meissner-Bernard, Claire and Titze, Benjamin and Chung, SueYeon and Friedrich, Rainer W},
  title = {Representational learning by optimization of neural manifolds in an olfactory memory network},
	elocation-id = {2024.11.17.623906},
	year = {2024},
	doi = {10.1101/2024.11.17.623906},
  publisher = {Cold Spring Harbor Laboratory},
  journal = {bioRxiv}
}

Higher brain functions depend on experience-dependent representations of relevant information that may be organized by attractor dynamics or by geometrical modifications of continuous "neural manifolds". To explore these scenarios we analyzed odor-evoked activity in telencephalic area pDp of juvenile and adult zebrafish, the homolog of piriform cortex. No obvious signatures of attractor dynamics were detected. Rather, olfactory discrimination training selectively enhanced the separation of neural manifolds representing task-relevant odors from other representations, consistent with predictions of autoassociative network models endowed with precise synaptic balance. Analytical approaches using the framework of manifold capacity revealed multiple geometrical modifications of representational manifolds that supported the classification of task-relevant sensory information. Manifold capacity predicted odor discrimination across individuals, indicating a close link between manifold geometry and behavior. Hence, pDp and possibly related recurrent networks store information in the geometry of representational manifolds, resulting in joint sensory and semantic maps that may support distributed learning processes.Competing Interest StatementThe authors have declared no competing interest.

@article{MCC24,
  title={Nonlinear classification of neural manifolds with contextual information},
  author={Francesca Mignacco and Chi-Ning Chou and SueYeon Chung},
  year={2024},
  eprint={2405.06851},
  archivePrefix={arXiv},
  primaryClass={q-bio.NC}
}

Understanding how neural systems efficiently process information through distributed representations is a fundamental challenge at the interface of neuroscience and machine learning. Recent approaches analyze the statistical and geometrical attributes of neural representations as population-level mechanistic descriptors of task implementation. In particular, manifold capacity has emerged as a promising framework linking population geometry to the separability of neural manifolds. However, this metric has been limited to linear readouts. Here, we propose a theoretical framework that overcomes this limitation by leveraging contextual input information. We derive an exact formula for the context-dependent capacity that depends on manifold geometry and context correlations, and validate it on synthetic and real data. Our framework's increased expressivity captures representation untanglement in deep networks at early stages of the layer hierarchy, previously inaccessible to analysis. As context-dependent nonlinearity is ubiquitous in neural systems, our data-driven and theoretically grounded approach promises to elucidate context-dependent computation across scales, datasets, and models.

@article {Chou2024.02.26.582157,
  author = {Chi-Ning Chou and Luke Arend and Albert J Wakhloo and Royoung Kim and Will Slatton and SueYeon Chung},
  title = {Neural Manifold Capacity Captures Representation Geometry, Correlations, and Task-Efficiency Across Species and Behaviors},
  elocation-id = {2024.02.26.582157},
  year = {2024},
  doi = {10.1101/2024.02.26.582157},
  publisher = {Cold Spring Harbor Laboratory},
  journal = {bioRxiv}
}

The study of the brain encompasses multiple scales, including temporal, spatial, and functional aspects. To integrate understanding across these different levels and modalities, it requires developing quantification methods and frameworks. Here, we present effective Geometric measures from Correlated Manifold Capacity theory (GCMC) for probing the functional structure in neural representations. We utilize a statistical physics approach to establish analytical connections between neural co-variabilities and downstream read-out efficiency. These effective geometric measures capture both stimulus-driven and behavior-driven structures in neural population activities, while extracting computationally-relevant information from neural data into intuitive and interpretable analysis descriptors. We apply GCMC to a diverse collection of datasets with different recording methods, various model organisms, and multiple task modalities. Specifically, we demonstrate that GCMC enables a wide range of multi-scale data analysis. This includes quantifying the spatial progression of encoding efficiency across brain regions, revealing the temporal dynamics of task-relevant manifold geometry in information processing, and characterizing variances as well as invariances in neural representations throughout learning. Lastly, the effective manifold geometric measures may be viewed as order parameters for phases related to computational efficiency, facilitating data-driven hypothesis generation and latent embedding.

@article{YCC22,
  title={Understanding Rare Spurious Correlations in Neural Networks},
  author={Yang, Yao-Yuan and Chou, Chi-Ning and Chaudhuri, Kamalika},
  eprint={2202.05189},
  archivePrefix={arXiv},
  primaryClass={cs.LG}
}

Neural networks are known to use spurious correlations such as background information for classification. While prior work has looked at spurious correlations that are widespread in the training data, in this work, we investigate how sensitive neural networks are to rare spurious correlations, which may be harder to detect and correct, and may lead to privacy leaks. We introduce spurious patterns correlated with a fixed class to a few training examples and find that it takes only a handful of such examples for the network to learn the correlation. Furthermore, these rare spurious correlations also impact accuracy and privacy. We empirically and theoretically analyze different factors involved in rare spurious correlations and propose mitigation methods accordingly. Specifically, we observe that $\ell_2$ regularization and adding Gaussian noise to inputs can reduce the undesirable effects. Code available at [this https URL](https://github.com/yangarbiter/rare-spurious-correlation){:target="_blank"}.

@article{CGSV21,
  title={Approximability of all Boolean CSPs in the dynamic streaming setting},
  author={Chou, Chi-Ning and Golovnev, Sasha and Sudan, Madhu and Velusamy, Santhoshini},
  year={2021},
  archivePrefix = "arXiv",
  eprint = "2102.12351",
  primaryClass={cs.CC}
}

A Boolean constraint satisfaction problem (CSP), $\textsf{Max-CSP}(f)$, is a maximization problem specified by a constraint $f:\{-1,1\}^k\to\{0,1\}$. An instance of the problem consists of $m$ constraint applications on $n$ Boolean variables, where each constraint application applies the constraint to $k$ literals chosen from the $n$ variables and their negations. The goal is to compute the maximum number of constraints that can be satisfied by a Boolean assignment to the $n$~variables. In the $(\gamma,\beta)$-approximation version of the problem for parameters $\gamma \geq \beta \in [0,1]$, the goal is to distinguish instances where at least $\gamma$ fraction of the constraints can be satisfied from instances where at most $\beta$ fraction of the constraints can be satisfied.

In this work we consider the approximability of $\textsf{Max-CSP}(f)$ in the (dynamic) streaming setting, where constraints are inserted (and may also be deleted in the dynamic setting) one at a time. We completely characterize the approximability of all Boolean CSPs in the dynamic streaming setting. Specifically, given $f$, $\gamma$ and $\beta$ we show that either (1) the $(\gamma,\beta)$-approximation version of $\textsf{Max-CSP}(f)$ has a probabilistic dynamic streaming algorithm using $O(\log n)$ space, or (2) for every $\epsilon > 0$ the $(\gamma-\epsilon,\beta+\epsilon)$-approximation version of $\textsf{Max-CSP}(f)$ requires $\Omega(\sqrt{n})$ space for probabilistic dynamic streaming algorithms. We also extend previously known results in the insertion-only setting to a wide variety of cases, and in particular the case of $k=2$ where we get a dichotomy and the case when the satisfying assignments of $f$ support a distribution on $\{-1,1\}^k$ with uniform marginals.

Our positive results show wider applicability of bias-based algorithms used previously by [Guruswami-Velingker-Velusamy, APPROX 2017]] and [Chou-Golovnev-Velusamy, FOCS 2020] by giving a systematic way to discover biases. Our negative results combine the Fourier analytic methods of [Kapralov-Khannan-Sudan, SODA 2015], which we extend to a wider class of CSPs, with a rich collection of reductions among communication complexity problems that lie at the heart of the negative results.

@article{CSWY21,
  title={A General Framework for Analyzing Stochastic Dynamics in Learning Algorithms},
  author={Chou, Chi-Ning and Sandhu, Juspreet Singh and Wang, Mien Brabeeba and Yu, Tiancheng},
  eprint={2006.06171},
  archivePrefix={arXiv},
  primaryClass={math.OC}
}

One of the challenges in analyzing a learning algorithm is the circular entanglement between the objective value and the stochastic noise. This is also known as the "chicken and egg" phenomenon. Traditionally, people tackle this issue with the special structure of the problem and hence the analysis is difficult to generalize.

In this paper, we present a general framework for analyzing high-probability bounds for stochastic dynamics in learning algorithms. Our framework composes standard techniques from probability theory to give a streamlined three-step recipe with a general and flexible principle to tackle the "chicken and egg" problem. We demonstrate the power and the flexibility of our framework by giving unifying analysis for three very different learning problems with both the last iterate and the strong uniform high probability convergence guarantee. The problems are stochastic gradient descent for strongly convex functions, streaming principal component analysis and linear bandit with stochastic gradient descent updates. We either improve or match the state-of-the-art bounds on all three dynamics.

@unpublished{CKS19,
  title={Closure of VP under taking factors: a short and simple proof},
  author={Chou, Chi-Ning and Kumar, Mrinal and Solomon, Noam},
  note={In submission to Theory of Computing},
  eprint={1903.02366},
  archivePrefix={arXiv},
  primaryClass={cs.CC}
}

In this note, we give a short, simple and almost completely self contained proof of a classical result of Kaltofen which shows that if an $n$ variate degree $d$ polynomial $f$ can be computed by an arithmetic circuit of size $s$, then each of its factors can be computed by an arithmetic circuit of size at most $\poly\left(s, n, d\right)$.

However, unlike Kaltofen's argument, our proof does not directly give an efficient algorithm for computing the circuits for the factors of $f$.