论文检索

Deep Learning · Algorithms

Chi Zhang, Jingpu Cheng, Yanyu Xu, Qianxiao Li

In contrast to the prevailing interpretation of Low-Rank Adaptation (LoRA) as a means of simulating weight changes in model adaptation, this paper introduces an alternative perspective by framing it as a control process. Specifically, we conceptualize lightweight matrices in LoRA as control modules tasked with perturbing the original, complex, yet frozen blocks on downstream tasks. Building upon this new understanding, we conduct a thorough analysis on the controllability of these modules, where we identify and establish sufficient conditions that facilitate their effective integration into downstream controls. Moreover, the control modules are redesigned by incorporating nonlinearities through a parameter-free attention mechanism. This modification allows for the intermingling of tokens within the controllers, enhancing the adaptability and performance of the system. Empirical findings substantiate that, without introducing any additional parameters, this approach surpasses the existing LoRA algorithms across all assessed datasets and rank configurations.

Applications

Chen Zhang, Qiang HE, Yuan Zhou, Elvis S. Liu, Hong Wang, Jian Zhao, Yang Wang

Deep Reinforcement Learning (DRL) agents have demonstrated impressive success in a wide range of game genres. However, existing research primarily focuses on optimizing DRL competence rather than addressing the challenge of prolonged player interaction. In this paper, we propose a practical DRL agent system for fighting games named _Shūkai_, which has been successfully deployed to Naruto Mobile, a popular fighting game with over 100 million registered users. _Shūkai_ quantifies the state to enhance generalizability, introducing Heterogeneous League Training (HELT) to achieve balanced competence, generalizability, and training efficiency. Furthermore, _Shūkai_ implements specific rewards to align the agent's behavior with human expectations. _Shūkai_'s ability to generalize is demonstrated by its consistent competence across all characters, even though it was trained on only 13% of them. Additionally, HELT exhibits a remarkable 22% improvement in sample efficiency. _Shūkai_ serves as a valuable training partner for players in Naruto Mobile, enabling them to enhance their abilities and skills.

Theory · Reinforcement Learning and Planning

Kaiwen Wang, Owen Oertell, Alekh Agarwal, Nathan Kallus, Wen Sun

In this paper, we prove that Distributional Reinforcement Learning (DistRL), which learns the return distribution, can obtain second-order bounds in both online and offline RL in general settings with function approximation. Second-order bounds are instance-dependent bounds that scale with the variance of return, which we prove are tighter than the previously known small-loss bounds of distributional RL. To the best of our knowledge, our results are the first second-order bounds for low-rank MDPs and for offline RL. When specializing to contextual bandits (one-step RL problem), we show that a distributional learning based optimism algorithm achieves a second-order worst-case regret bound, and a second-order gap dependent bound, simultaneously. We also empirically demonstrate the benefit of DistRL in contextual bandits on real-world datasets. We highlight that our analysis with DistRL is relatively simple, follows the general framework of optimism in the face of uncertainty and does not require weighted regression. Our results suggest that DistRL is a promising framework for obtaining second-order bounds in general RL settings, thus further reinforcing the benefits of DistRL.

Theory · Reinforcement Learning and Planning

Shuai Zhang, Heshan Fernando, Miao Liu, Keerthiram Murugesan, Songtao Lu, Pin-Yu Chen, Tianyi Chen, Meng Wang

This paper studies the transfer reinforcement learning (RL) problem where multiple RL problems have different reward functions but share the same underlying transition dynamics. In this setting, the Q-function of each RL problem (task) can be decomposed into a successor feature (SF) and a reward mapping: the former characterizes the transition dynamics, and the latter characterizes the task-specific reward function. This Q-function decomposition, coupled with a policy improvement operator known as generalized policy improvement (GPI), reduces the sample complexity of finding the optimal Q-function, and thus the SF & GPI framework exhibits promising empirical performance compared to traditional RL methods like Q-learning. However, its theoretical foundations remain largely unestablished, especially when learning the successor features using deep neural networks (SF-DQN). This paper studies the provable knowledge transfer using SFs-DQN in transfer RL problems. We establish the first convergence analysis with provable generalization guarantees for SF-DQN with GPI. The theory reveals that SF-DQN with GPI outperforms conventional RL approaches, such as deep Q-network, in terms of both faster convergence rate and better generalization. Numerical experiments on real and synthetic RL tasks support the superior performance of SF-DQN & GPI, aligning with our theoretical findings.

Applications · Physics

Haoyu Yang, Mark Ren

Lithography, transferring chip design masks to the silicon wafer, is the most important phase in modern semiconductor manufacturing flow. Due to the limitations of lithography systems, Extensive design optimizations are required to tackle the design and silicon mismatch. Inverse lithography technology (ILT) is one of the promising solutions to perform pre-fabrication optimization, termed mask optimization. Because of mask optimization problems’ constrained non-convexity, numerical ILT solvers rely heavily on good initialization to avoid getting stuck on sub-optimal solutions. Machine learning (ML) techniques are hence proposed to generate mask initialization for ILT solvers with one-shot inference, targeting faster and better convergence during ILT. This paper addresses the question of whether ML models can directly generate high-quality optimized masks without engaging ILT solvers in the loop. We propose an implicit learning ILT framework: ILILT, which leverages the implicit layer learning method and lithography-conditioned inputs to ground the model. Trained to understand the ILT optimization procedure, ILILT can outperform the state-of-the-art machine learning solutions, significantly improving efficiency and quality.

Deep Learning · Generative Models and Autoencoders

Yahong Zhang, Chenchen Fan, Donghui Chen, Congrui Li, Wenli Ouyang, Mingda Zhu, Junchi Yan

Machine learning (ML) has been actively adopted in Linear Programming (LP) and Mixed-Integer Linear Programming (MILP), whose potential is hindered by instance scarcity. Current synthetic instance generation methods often fall short in closely mirroring the distribution of original datasets or ensuring the feasibility and boundedness of the generated data — a critical requirement for obtaining reliable supervised labels in model training. In this paper, we present a diffusion-based LP/MILP instance generative framework called MILP-FBGen. It strikes a balance between structural similarity and novelty while maintaining feasibility/boundedness via a meticulously designed structure-preserving generation module and a feasibility/boundedness-constrained sampling module. Our method shows superiority on two fronts: 1) preservation of key properties (hardness, feasibility, and boundedness) of LP/MILP instances, and 2) enhanced performance on downstream tasks. Extensive studies show two-fold superiority that our method ensures higher distributional similarity and 100% feasibility in both easy and hard datasets, surpassing current state-of-the-art techniques.

Social Aspects · Trustworthy Machine Learning

Hanlin Zhang, Benjamin Edelman, Danilo Francati, Daniele Venturi, Giuseppe Ateniese, Boaz Barak

Watermarking generative models consists of planting a statistical signal (watermark) in a model's output so that it can be later verified that the output was generated by the given model. A strong watermarking scheme satisfies the property that a computationally bounded attacker cannot erase the watermark without causing significant quality degradation. In this paper, we study the (im)possibility of strong watermarking schemes. We prove that, under well-specified and natural assumptions, strong watermarking is impossible to achieve. This holds even in the private detection algorithm setting, where the watermark insertion and detection algorithms share a secret key, unknown to the attacker. To prove this result, we introduce a generic efficient watermark attack; the attacker is not required to know the private key of the scheme or even which scheme is used. Our attack is based on two assumptions: (1) The attacker has access to a "quality oracle" that can evaluate whether a candidate output is a high-quality response to a prompt, and (2) The attacker has access to a "perturbation oracle" which can modify an output with a nontrivial probability of maintaining quality, and which induces an efficiently mixing random walk on high-quality outputs. We argue that both assumptions can be satisfied in practice by an attacker with weaker computational capabilities than the watermarked model itself, to which the attacker has only black-box access. Furthermore, our assumptions will likely only be easier to satisfy over time as models grow in capabilities and modalities. We demonstrate the feasibility of our attack by instantiating it to attack three existing watermarking schemes for large language models: Kirchenbauer et al. (2023), Kuditipudi et al. (2023), and Zhao et al. (2023), and include preliminary results on vision-language models. The same attack successfully removes the watermarks planted by all schemes, with only minor quality degradation.

Deep Learning · Large Language Models

Yuxin Zhang, Yuxuan Du, Gen Luo, Yunshan Zhong, Zhenyu Zhang, Shiwei Liu, Rongrong Ji

Despite the exceptional performance of Large Language Models (LLMs), the substantial volume of key-value (KV) pairs cached during inference presents a barrier to their efficient deployment. To ameliorate this, recent works have aimed to selectively eliminate these caches, informed by the attention scores of associated tokens. However, such cache eviction invariably leads to output perturbation, regardless of the token choice. This perturbation escalates with the compression ratio, which can precipitate a marked deterioration in LLM inference performance. This paper introduces Cache Merging (CaM) as a solution to mitigate this challenge. CaM adaptively merges to-be-evicted caches into the remaining ones, employing a novel sampling strategy governed by the prominence of attention scores within discarded locations. In this manner, CaM enables memory-efficient LLMs to preserve critical token information, even obviating the need to maintain their corresponding caches. Extensive experiments utilizing LLaMA, OPT, and GPT-NeoX across various benchmarks corroborate CaM's proficiency in bolstering the performance of memory-efficient LLMs. Code is released at https://github.com/zyxxmu/cam.

Deep Learning · Self-Supervised Learning

Qi Zhang, Tianqi Du, Haotian Huang, Yifei Wang, Yisen Wang

In recent years, the rise of generative self-supervised learning (SSL) paradigms has exhibited impressive performance across visual, language, and multi-modal domains. While the varied designs of generative SSL objectives lead to distinct properties in downstream tasks, a theoretical understanding of these differences remains largely unexplored. In this paper, we establish the first theoretical comparisons between two leading generative SSL paradigms: autoregressive SSL and masked SSL. Through establishing theoretical frameworks, we elucidate the strengths and limitations of autoregressive and masked SSL within the primary evaluation tasks of classification and content generation. Our findings demonstrate that in classification tasks, the flexibility of targeted tokens in masked SSL fosters more inter-sample connections compared to the fixed position of target tokens in autoregressive SSL, which yields superior clustering performance. In content generation tasks, the misalignment between the flexible lengths of test samples and the fixed length of unmasked texts in masked SSL (vs. flexible lengths of conditional texts in autoregressive SSL) hinders its generation performance. To leverage each other's strengths and mitigate weaknesses, we propose diversity-enhanced autoregressive and variable-length masked objectives, which substantially improve the classification performance of autoregressive SSL and the generation performance of masked SSL. Code is available at https://github.com/PKU-ML/LookAheadLookAround.

Optimization · Non-Convex

Qinzi Zhang, Ashok Cutkosky

Training neural networks requires optimizing a loss function that may be highly irregular, and in particular neither convex nor smooth. Popular training algorithms are based on stochastic gradient descent with momentum (SGDM), for which classical analysis applies only if the loss is either convex or smooth. We show that a very small modification to SGDM closes this gap: simply scale the update at each time point by an exponentially distributed random scalar. The resulting algorithm achieves optimal convergence guarantees. Intriguingly, this result is not derived by a specific analysis of SGDM: instead, it falls naturally out of a more general framework for converting online convex optimization algorithms to non-convex optimization algorithms.

Optimization · Everything Else

Cheng Zhang, Jianyi Cheng, George Constantinides, Yiren Zhao

Post-training quantization of Large Language Models (LLMs) is challenging. In this work, we introduce **L**ow-rank **Q**uantization **E**rror **R**eduction (LQER), which combines quantization and low-rank approximation to recover the model capability. LQER leverages an activation-induced scale matrix to drive the singular value distribution of quantization error towards a desirable distribution, which enables nearly-lossless W4A8 quantization on various LLMs and downstream tasks without the need for knowledge distillation, grid search, or gradient-based iterative optimization. Unlike existing methods, the computation pattern of LQER eliminates the need for specialized Scatter and Gather processes to collect high-precision weights from irregular memory locations. Our W4A8 LLMs achieve near-lossless performance on six popular downstream tasks, while using $1.36 \times$ fewer hardware resources than the leading state-of-the-art method. We will open-source our framework at [https://github.com/ChengZhang-98/lqer](https://github.com/ChengZhang-98/lqer)

Applications · Physics

Sergei Shumilin, Alexander Ryabov, Nikolay Yavich, Evgeny Burnaev, Vladimir Vanovskiy

Due to the high computational load of modern numerical simulation, there is a demand for approaches that would reduce the size of discrete problems while keeping the accuracy reasonable. In this work, we present an original algorithm to coarsen an unstructured grid based on the concepts of differentiable physics. We achieve this by employing $k$-means clustering, autodifferentiation and stochastic minimization algorithms. We demonstrate performance of the designed algorithm on two PDEs: a linear parabolic equation which governs slightly compressible fluid flow in porous media and the wave equation. Our results show that in the considered scenarios, we reduced the number of grid points up to 10 times while preserving the modeled variable dynamics in the points of interest. The proposed approach can be applied to the simulation of an arbitrary system described by evolutionary partial differential equations.

Theory · Reinforcement Learning and Planning

Tonghe Zhang, Yu Chen, Longbo Huang

This work pioneers regret analysis of risk-sensitive reinforcement learning in partially observable environments with hindsight observation, addressing a gap in theoretical exploration. We introduce a novel formulation that integrates hindsight observations into a Partially Observable Markov Decision Process (POMDP) framework, where the goal is to optimize accumulated reward under the entropic risk measure. We develop the first provably efficient RL algorithm tailored for this setting. We also prove by rigorous analysis that our algorithm achieves polynomial regret $\tilde{O}\left(\frac{e^{|{\gamma}|H}-1}{|{\gamma}|H}H^2\sqrt{KHS^2OA}\right)$, which outperforms or matches existing upper bounds when the model degenerates to risk-neutral or fully observable settings. We adopt the method of change-of-measure and develop a novel analytical tool of beta vectors to streamline mathematical derivations. These techniques are of particular interest to the theoretical study of reinforcement learning.

Miscellaneous Aspects of Machine Learning · Online Learning, Active Learning and Bandits

Zhiyu Zhang, David Bombara, Heng Yang

We study online learning in adversarial nonstationary environments. Since the future can be very different from the past, a critical challenge is to gracefully forget the history while new data comes in. To formalize this intuition, we revisit the discounted regret in online convex optimization, and propose an adaptive (i.e., instance optimal), FTRL-based algorithm that improves the widespread non-adaptive baseline -- gradient descent with a constant learning rate. From a practical perspective, this refines the classical idea of regularization in lifelong learning: we show that designing better regularizers can be guided by the principled theory of adaptive online optimization. Complementing this result, we also consider the (Gibbs & Candes, 2021)-style online conformal prediction problem, where the goal is to sequentially predict the uncertainty sets of a black-box machine learning model. We show that the FTRL nature of our algorithm can simplify the conventional gradient-descent-based analysis, leading to instance-dependent performance guarantees.

Optimization · Stochastic

Lijun Zhang, Haomin Bai, Wei-Wei Tu, Ping Yang, Yao Hu

While traditional distributionally robust optimization (DRO) aims to minimize the maximal risk over a set of distributions, Agarwal & Zhang (2022) recently proposed a variant that replaces risk with *excess risk*. Compared to DRO, the new formulation—minimax excess risk optimization (MERO) has the advantage of suppressing the effect of heterogeneous noise in different distributions. However, the choice of excess risk leads to a very challenging minimax optimization problem, and currently there exists only an inefficient algorithm for empirical MERO. In this paper, we develop efficient stochastic approximation approaches which directly target MERO. Specifically, we leverage techniques from stochastic convex optimization to estimate the minimal risk of every distribution, and solve MERO as a stochastic convex-concave optimization (SCCO) problem with biased gradients. The presence of bias makes existing theoretical guarantees of SCCO inapplicable, and fortunately, we demonstrate that the bias, caused by the estimation error of the minimal risk, is under-control. Thus, MERO can still be optimized with a nearly optimal convergence rate. Moreover, we investigate a practical scenario where the quantity of samples drawn from each distribution may differ, and propose a stochastic approach that delivers *distribution-dependent* convergence rates.

Theory · Game Theory

Youzhi Zhang, Bo An, Daniel Zeng

Many works recently have focused on computing optimal solutions for the ex ante coordination of a team for solving sequential adversarial team games, where a team of players coordinate against an opponent (or a team of players) in a zero-sum extensive-form game. However, it is challenging to directly compute such an optimal solution because the team’s coordinated strategy space is exponential in the size of the game tree due to the asymmetric information of team members. Column Generation (CG) algorithms have been proposed to overcome this challenge by iteratively expanding the team’s coordinated strategy space via a Best Response Oracle (BRO). More recently, more compact representations (particularly, the Team Belief Directed Acyclic Graph (TB-DAG)) of the team’s coordinated strategy space have been proposed, but the TB-DAG-based algorithms only outperform the CG-based algorithms in games with a small TB-DAG. Unfortunately, it is inefficient to directly apply CG to the TB-DAG because the size of the TB-DAG is still exponential in the size of the game tree and then makes the BRO unscalable. To this end, we develop our novel TB-DAG CG (DCG) algorithm framework by computing a coordinated best response in the original game first and then transforming this strategy into the TB-DAG form. To further improve the scalability, we propose a more suitable BRO for DCG to reduce the cost of the transformation at each iteration. We theoretically show that our algorithm converges exponentially faster than the state-of-the-art CG algorithms, and experimental results show that our algorithm is at least two orders of magnitude faster than the state-of-the-art baselines.

Miscellaneous Aspects of Machine Learning · Causality

Zhiheng Zhang

Partial identification (PI) presents a significant challenge in causal inference due to the incomplete measurement of confounders. Given that obtaining auxiliary variables of confounders is not always feasible and relies on untestable assumptions, researchers are encouraged to explore the internal information of latent confounders without external assistance. However, these prevailing PI results often lack precise mathematical measurement from observational data or assume that the information pertaining to confounders falls within extreme scenarios. In our paper, we reassess the significance of the marginal confounder distribution in PI. We refrain from imposing additional restrictions on the marginal confounder distribution, such as entropy or mutual information. Instead, we establish the closed-form tight PI for any possible P(U) in the discrete case. Furthermore, we establish the if and only if criterion for discerning whether the marginal confounder information leads to non-vanilla PI regions. This reveals a fundamental negative result wherein the marginal confounder information minimally contributes to PI as the confounder’s cardinality increases. Our theoretical findings are supported by experiments.

Optimization · Optimization and Learning under Uncertainty

Ali Zeynali, Shahin Kamali, Mohammad Hajiesmaili

We present the first learning-augmented data structure for implementing dictionaries with optimal consistency and robustness. Our data structure, named RobustSL, is a Skip list augmented by predictions of access frequencies of elements in a data sequence. With proper predictions, RobustSL has optimal consistency (achieves static optimality). At the same time, it maintains a logarithmic running time for each operation, ensuring optimal robustness, even if predictions are generated adversarially. Therefore, RobustSL has all the advantages of the recent learning-augmented data structures of Lin, Luo, and Woodruff (ICML 2022) and Cao et al. (arXiv 2023), while providing robustness guarantees that are absent in the previous work. Numerical experiments show that RobustSL outperforms alternative data structures using both synthetic and real datasets.

Probabilistic Methods · Variational Inference

Johannes Zenn, Robert Bamler

Differentiable annealed importance sampling (DAIS), proposed by Geffner & Domke (2021) and Zhang et al. (2021), allows optimizing, among others, over the initial distribution of AIS. In this paper, we show that, in the limit of many transitions, DAIS minimizes the symmetrized KL divergence (Jensen-Shannon divergence) between the initial and target distribution. Thus, DAIS can be seen as a form of variational inference (VI) in that its initial distribution is a parametric fit to an intractable target distribution. We empirically evaluate the usefulness of the initial distribution as a variational distribution on synthetic and real-world data, observing that it often provides more accurate uncertainty estimates than standard VI (optimizing the reverse KL divergence), importance weighted VI, and Markovian score climbing (optimizing the forward KL divergence).

Deep Learning · Algorithms

Zhanpeng Zeng, Karthikeyan Sankaralingam, Vikas Singh

GEneral Matrix Multiply (GEMM) is a central operation in deep learning and corresponds to a large chunk of the compute footprint. Therefore, improving its efficiency is an active topic of research. A popular strategy is the use of low bit-width integers to approximate the original matrix entries. This allows efficiency gains, but often requires sophisticated techniques to control the rounding error. In this work, we first verify that when the low bit-width restriction is removed, for a variety of Transformer-based models, integers are, in fact, sufficient for all GEMMs need -- for both training and inference stages, and achieve parity (with floating point). No sophisticated techniques are needed. We find that while a large majority of entries in matrices (encountered in such models) can be easily represented by low bit-width integers, the existence of a few heavy hitter entries make it difficult to achieve efficiency gains via the exclusive use of low bit-width GEMMs alone. To address this issue, we develop a simple algorithm, Integer Matrix Unpacking (IM-Unpack), to unpack a matrix with large integer entries into a larger matrix whose entries all lie within the representable range of arbitrarily low bit-width integers. This allows equivalence with the original GEMM, i.e., the exact result can be obtained using purely low bit-width integer GEMMs. This comes at the cost of additional operations -- we show that for many popular models, this overhead is quite small. Code is available at https://github.com/vsingh-group/im-unpack.