论文检索

Deep Learning · Large Language Models

Jie Hao, Rui Yu, Wei Zhang, Huixia Judy Wang, Jie Xu, Mingrui Liu

Effective data selection is essential for pretraining large language models (LLMs), enhancing efficiency and improving generalization to downstream tasks. However, existing approaches often require leveraging external pretrained models, making it difficult to disentangle the effects of data selection from those of the external pretrained models. In addition, they often overlook the long-term impact of selected data if the model is trained for a long period of time, primarily due to the prohibitive cost of full-scale LLM pretraining. In this paper, we introduce BLISS (**B**ileve**L** **I**nfluence **S**coring method for data **S**election): a lightweight data selection method that operates entirely \emph{from scratch}, without relying on any external pretrained oracle models, while explicitly accounting for the long-term impact of selected data. BLISS leverages a small proxy model as a surrogate for the LLM and employs a score model to estimate the long-term influence of training samples if the proxy model is trained to convergence. We formulate data selection as a bilevel optimization problem, where the upper-level objective optimizes the score model to assign importance weights to training samples, ensuring that minimizing the lower-level objective (i.e., training the proxy model over the weighted training loss until convergence) leads to best validation performance. Once optimized, the trained score model predicts influence scores for the dataset, enabling efficient selection of high-quality samples for LLM pretraining. We validate BLISS by pretraining 410M/1B/2.8B Pythia and LLaMA-0.5B models on selected subsets of the C4 dataset. Notably, under the 1B model setting, BLISS achieves $1.7\times$ speedup in reaching the same performance as the state-of-the-art method, demonstrating superior performance across multiple downstream tasks.

Haoyu Wang, Haiyan Zhao, Xingyu Yu, Zhangyang Yao, Xu Han, Zhiyuan Liu, Maosong Sun

Post-training quantization at the 2-bit level enables low-cost deployment and inference acceleration for large language models (LLMs). Scalar quantization (SQ) and vector quantization (VQ) are two primary quantization methods, however, the former suffers from significant performance degradation, and the latter incurs computational and storage overhead. We propose UniSVQ, a unified 2-bit quantization framework that bridges scalar and vector quantization by parameterizing codewords as an affine transform of integer lattices. This structure preserves compatibility with optimized integer kernels while retaining much of VQ's flexibility. We further introduce a data-driven block-wise fine-tuning strategy to directly minimize quantization reconstruction error. Extensive experiments across multiple LLM families and zero-shot benchmarks demonstrate that UniSVQ consistently outperforms state-of-the-art SQ methods and achieves performance comparable to advanced VQ methods, while providing higher inference throughput.

Darshan Deshpande, Anand Kannappan, Rebecca Qian

Recent advances in reinforcement learning for code generation have made robust environments essential to prevent reward hacking. As LLMs increasingly serve as evaluators in code-based RL, their ability to detect reward hacking remains understudied. In this paper, we propose a novel taxonomy of reward exploits spanning across 54 categories and introduce TRACE (Testing Reward Anomalies in Code Environments), a synthetically curated and human-verified benchmark containing 517 testing trajectories. Unlike prior work that evaluates reward hack detection in isolated classification scenarios, we contrast these evaluations with a more realistic, contrastive anomaly detection setup on TRACE. Our experiments reveal that models capture reward hacks more effectively in contrastive settings than in isolated classification settings, with GPT-5.2 with highest reasoning mode achieving the best detection rate at 63%, up from 45% in isolated settings on TRACE. Building on this insight, we demonstrate that state-of-the-art models struggle significantly more with semantically contextualized reward hacks compared to syntactically contextualized ones. We further conduct qualitative analyses of model behaviors, as well as ablation studies showing that the ratio of benign to hacked trajectories and analysis cluster sizes substantially impact detection performance. We release the benchmark and evaluation harness to enable the community to expand TRACE and evaluate their models.

Reinforcement Learning · Everything Else

Yang Shengtian, Ziteng Cui, Shuo He, Yewen Li, Qingpeng Cai, Peng Jiang, Lei Feng

Reinforcement learning (RL) has equipped LLM agents with a strong ability to solve complex tasks. However, existing RL methods normally use a single policy network, causing simplicity bias where simple tasks occupy most parameters and dominate gradient updates, leaving insufficient capacity for complex tasks. A plausible remedy could be employing the Mixture-of-Experts (MoE) architecture in the policy network, as MoE allows different parameters (experts) to specialize in different tasks, preventing simple tasks from dominating all parameters. However, a key limitation of traditional MoE is its token-level routing, where the router assigns each token to specialized experts, which fragments phase-consistent patterns into scattered expert assignments and thus undermines expert specialization. In this paper, we propose Phase-Aware Mixture of Experts (PA-MoE). It first features a lightweight phase router that learns latent phase boundaries directly from the RL objective without pre-defining phase categories. Then, the phase router allocates temporally consistent assignments to the same expert, allowing experts to preserve phase-specific expertise. Experimental results demonstrate the effectiveness of our proposed PA-MoE. Code is available at https://anonymous.4open.science/r/PA-MoE-576C/.

Applications · Chemistry, Physics, and Earth Sciences

Shreshth Malik, Tiarnan Doherty, Panagiotis Tigas, Muhammed Razzak, Stephen Roberts, Aron Walsh, Yarin Gal

Existing benchmarks for computational materials discovery primarily evaluate static predictive tasks or isolated computational sub-tasks. While valuable, these evaluations neglect the inherently iterative and adaptive nature of scientific discovery. We introduce MAterials Discovery Environments (MADE), a novel framework for benchmarking end-to-end autonomous materials discovery pipelines. MADE simulates closed-loop discovery campaigns in which an agent or algorithm proposes, evaluates, and refines candidate materials under a constrained oracle budget, capturing the sequential and resource-limited nature of real discovery workflows. We formalize discovery as a search for thermodynamically stable compounds relative to a given convex hull, and evaluate efficacy and efficiency via comparison to baseline algorithms. The framework is flexible; users can compose discovery agents from interchangeable components such as generative models, filters, and planners, enabling the study of arbitrary workflows ranging from fixed pipelines to fully agentic systems with tool use and adaptive decision making. We demonstrate this by conducting systematic experiments across a family of systems, enabling ablation of components in discovery pipelines, and comparison of how methods scale with system complexity.

Deep Learning · Generative Models and Autoencoders

Sangeek Hyun, MinKyu Lee, Jae-Pil Heo

Scalability has driven recent advances in generative modeling, yet it remains underexplored for adversarial learning. We study the scaling behavior of Generative Adversarial Networks through two design choices: training in a compact Variational Autoencoder latent space and using purely transformer-based generators and discriminators. While this setup is efficient and scales well with compute, naively scaling exposes failure modes; underutilization of early layers in the generator and increasing optimization instability. We address these issues with lightweight intermediate supervision and width-aware learning-rate adjustment. Our Generative Adversarial Transformers (GAT) train reliably from small (S) to extra-large (XL) model sizes, and GAT-XL model achieves state-of-the-art single-step class-conditional generation on ImageNet at 256×256 resolution (FID of 2.18) in 60 epochs, requiring 4x fewer epochs than strong baselines.

Social Aspects · Security

Qing Wen, Haohao Li, Zhongjie Ba, Peng Cheng, Miao He, Li Lu, Kui Ren

Advances in AIGC technologies have enabled the synthesis of highly realistic audio deepfakes capable of deceiving human auditory perception. Although numerous audio deepfake detection (ADD) methods have been developed, most rely on local temporal/spectral features or pairwise relations, overlooking high-order interactions (HOIs). HOIs capture discriminative patterns that emerge from multiple feature components beyond their individual contributions. We propose HyperPotter, a hypergraph-based framework that explicitly models these synergistic HOIs through clustering-based hyperedges with class-aware prototype initialization. Extensive experiments demonstrate that HyperPotter surpasses its baseline by an average relative gain of 22.15\% across 11 datasets and outperforms state-of-the-art methods by 13.96\% on 4 challenging cross-domain datasets, demonstrating superior generalization to diverse attacks and speakers.

Kieran Didi, Sarah Alamdari, Alex Lu, Bruce Wittmann, Kadina Johnston, Ava Amini, Ali Madani, Maya Czeneszew, Christian Dallago, Kevin Yang

Machine learning methods that predict protein fitness from sequence remain sensitive to changes in data distributions, limiting generalization across common conditions encountered in protein engineering. Practically, protein engineers are thus left wondering about the effective utility of ML tools. The FLIP benchmark established protocols for testing generalization under some domain shifts, but it was limited to measurements of stability, binding, and viral capsid viability. We introduce FLIP2, a protein fitness benchmark spanning seven new datasets, including enzymes, protein-protein interactions, and light-sensitive proteins, as well as splits that measure generalization relevant to real-world protein engineering campaigns. Evaluating a suite of benchmark models across these datasets and suites reveals that simpler models often matched or outperformed fine-tuned protein language models on \ourset, challenging the utility of existing transfer learning techniques. Provenance for all datasets has been recorded and we redistribute all data CC-BY 4.0 to facilitate continued progress.

Deep Learning · Large Language Models

Senyu Han, Yilu Cao, Kai Yu, Lu Chen

Large language model (LLM) exists a subset of attention heads that are highly responsible for long-context processing. Existing work has identified different long-context heads in models, but their detection methods mainly rely on model inference on actual long texts and do not analyze the inherent properties of the head parameters. In this paper, we use kernel methods to analyze static *frequency kernels* formed by different rotation frequency components of attention heads, and we design a Long-context Potential Score (LPS) to measure the potential of attention heads in processing long contexts. Kernels of heads with high LPS exhibit concentrated low-frequency energy and low effective rank, which allow them to effectively capture highly specialized information from distant contexts. Experiments and analysis on long-context tasks and model behaviors show that the LPS metrics can well reflect the actual capability of heads on long contexts. Furthermore, by simply amplifying low-frequency kernels of heads with high retrieval potential, we can further improve model's performance on long-context tasks. Our metrics and head enhancement methods are fully static and offline, and they can be quickly conducted under low-resource constraints.

Deep Learning · Large Language Models

Li, Yaming Guo, Shenghao Gao, Xinlong Chen, Zuhao Xu, Ying Sun, Chao Wang, Hui Xiong

While Large language models (LLMs) have strong abilities, they generally rely on fine-tuning to supplement downstream task-specific knowledge. Due to the prohibitive memory overhead of full fine-tuning (FT), existing parameter-efficient fine-tuning techniques, e.g., LoRA and Adapters, update parameters only in low-rank or restricted subspaces. However, they fail to approximate FT---the performative fine-tuner---and risk performance degradation in tough tasks. Therefore, we naturally raise a *Low-cost Full Fine-tuning* question: Can we approach standard full fine-tuning in theory, yet with much lower costs in practice? Our key insight is that performing selective updates at each step can, theoretically, recover FT asymptotically, while being cost-effective and ignoring no parameter direction. This motivates a new general fine-tuning paradigm (called *Think-Touch*): we first predict potentials of parameter groups (*think*) and then update only the selected (*touch*) in one step. Theoretically, we show that under a very weak sufficient condition---divergence of the cumulative coverage of the expected gradient norm---any selection strategy can converge in the full-parameter space to a stationary point at which the FT admits no further first-order improvement. Besides, we further derive the general convergence rate for our paradigm and identify a post-hoc greedy strategy that is rate-optimal. Unfortunately, this strategy cannot be directly applied in practice due to its reliance on full and accurate gradient information. Thus, we propose a bandit-based method to online approximate this ideal strategy in the long run with a rigorous regret guarantee. Extensive experimental results on various tasks demonstrate the potential of our paradigm, including much lower space overheads against FT and better performance than LoRAs.

Yuxin Zhang, Ju Fan, Meihao Fan, Shaolei Zhang, Xiaoyong Du

Advanced agents are increasingly demonstrating the potential to operate as autonomous engineers, creating a growing demand for evaluation benchmarks that capture the complexity of real-world development. Such environments typically involve both complex code and large-scale data (i.e., file system). However, existing benchmarks usually evaluate code-centric or data-centric capabilities in isolation, leaving a clear gap with real development scenarios. In this paper, we bridge this gap by introducing CoDA-Bench, the first benchmark to jointly evaluate code and data intelligence in a data-intensive environment. We construct a data-intensive Linux sandbox based on the Kaggle ecosystem (containing hundreds of datasets), where agents must actively explore complex file hierarchies to identify relevant resources and generate code for data-driven analytical tasks. CoDA-Bench comprises 1,202 tasks spanning 53 communities, with each task environment containing an average of 700 files, simulating realistic data scale and noise. Evaluations of advanced agents reveal that even top-performing systems struggle to effectively integrate data discovery with code execution, achieving a success rate of only 56.1%. These results highlight a substantial gap in current agentic capabilities for data-intensive tasks and point to promising directions for future research.

Kieran Didi, Sarah Alamdari, Alex Lu, Bruce Wittmann, Kadina Johnston, Ava Amini, Ali Madani, Maya Czeneszew, Christian Dallago, Kevin Yang

Machine learning methods that predict protein fitness from sequence remain sensitive to changes in data distributions, limiting generalization across common conditions encountered in protein engineering. Practically, protein engineers are thus left wondering about the effective utility of ML tools. The FLIP benchmark established protocols for testing generalization under some domain shifts, but it was limited to measurements of stability, binding, and viral capsid viability. We introduce FLIP2, a protein fitness benchmark spanning seven new datasets, including enzymes, protein-protein interactions, and light-sensitive proteins, as well as splits that measure generalization relevant to real-world protein engineering campaigns. Evaluating a suite of benchmark models across these datasets and suites reveals that simpler models often matched or outperformed fine-tuned protein language models on \ourset, challenging the utility of existing transfer learning techniques. Provenance for all datasets has been recorded and we redistribute all data CC-BY 4.0 to facilitate continued progress.

Shaolei Zhang, Ju Fan, Meihao Fan, Yizhe Liu, Yuxin Zhang, Xiaoyong Du

Autonomous data science on the structured data has been a long-standing challenge, and is now becoming feasible with the emergence of powerful large language models (LLMs). Recent workflowbased data agents have shown promising results on specific data tasks but remain fundamentally limited in achieving full autonomy due to their reliance on predefined workflows. In this paper, we introduce DeepAnalyze, the first agentic LLM for autonomous data science, capable of automatically completing the end-to-end data science from structured data to analyst-grade research reports. To tackle high-complexity data science tasks, we propose a curriculum-based agentic training paradigm that emulates the learning trajectory of human data scientists, enabling LLMs to progressively acquire and integrate multiple capabilities in real-world environments. Accordingly, we contribute a data-grounded trajectory synthesis framework to constructs high-quality data science training data. Through training in real-world environment, DeepAnalyze learns to perform a broad spectrum of data tasks, ranging from data question answering to open-ended data research. Experiments on 13 benchmarks demonstrate that, with only 8B parameters, DeepAnalyze outperforms previous workflow-based agents built on most advanced proprietary LLMs.

Applications · Everything Else

Hengnian Gu, Zhifu Chen, Yuxin Chen, Jin Zhou, Dongdai Zhou

Cognitive structure (CS), a student's construction of concepts and inter-concept relations, has long been recognized as a foundational notion in psychology and intelligent education, yet remains largely unassessable in practice. Existing approaches such as knowledge tracing (KT) and cognitive diagnosis (CD) simplify and indirectly approximate CS, but they intertwine representation learning with prediction objectives, limiting generalization, interpretability, and reuse across tasks. To address this gap, we propose Cognitive Structure Generation (CSG), a task-agnostic framework that explicitly models CS through generative modeling. Based on educational theories, CSG first pretrains a Cognitive Structure Diffusion Probabilistic Model (CSDPM) and then applies reinforcement learning with SOLO-based hierarchical rewards to capture plausible patterns of cognitive development. By decoupling cognitive structure representation from downstream prediction, CSG produces interpretable and transferable cognitive structures that can be seamlessly integrated into diverse student modeling tasks. Experiments on five real-world datasets show that CSG yields more comprehensive representations, substantially improving performance while offering enhanced interpretability and modularity.

Yuze Zhao, Kuiyuan Zhang, Zhongyun Hua, Yushu Zhang, Qing Liao, Wei Jiang

The rapid evolution of audio deepfakes requires robust detection capable of generalizing to unseen attacks. One-class learning offers inherent robustness for this task by characterizing real speech distributions to detect anomalies. However, establishing a compact decision boundary without spoof supervision remains a fundamental challenge. Existing relaxed approaches often compromise this strictness by introducing auxiliary negative samples, which biases the boundary toward seen artifacts and degrades generalization to unseen attacks. To address this, we propose CA-SOADD, a framework that refines the acceptance region by constructing off-manifold boundary probes. Our proposed centroid-anchored tri-objective learning paradigm simultaneously enforces centroid compactness and a centroid-referenced margin against these probes, thereby explicitly tightening the acceptance region without treating them as an explicit negative class. We further extend the framework to heterogeneous settings through domain-conditioned centroids. Experiments on ASVSpoof and MLAAD benchmarks demonstrate that our strict real-only method consistently outperforms strong baselines under unseen attack types and domain shifts, with its effectiveness further validated through extensive ablation studies.

Reinforcement Learning · Online

Duo Cheng, Xingyu Zhou, Bo Ji

We study hybrid Reinforcement Learning (RL) in adversarial Markov Decision Processes (MDPs), where the learner simultaneously receives on-policy feedback from the executed policy and off-policy feedback from a fixed behavior policy, and loss functions can change arbitrarily over time. On-policy feedback allows exploration and ensures the worst-case guarantee against any comparator policy, while off-policy feedback provides coverage-dependent guarantee that scales with the "mismatch" between the behavior and comparator policies (called coverage ratio) and can be sharper than on-policy results whenever this ratio is small. We propose a new hybrid RL framework that accommodates adversarial losses and unknown transitions, preserving off-policy guarantees while ensuring non-trivial worst-case performance.

Reinforcement Learning · Deep RL

Ruyi Lu, Xuesong Wang, Hengrui Zhang, Yuhu Cheng

Sample inefficiency remains a challenge in pixel-based visual reinforcement learning (RL), primarily due to ineffective state representation learning. While recent advances employ auxiliary tasks to improve representation learning, their representation goals (e.g., mask reconstruction, state prediction) are misaligned with the ultimate RL goal of maximizing return, constraining further improvements in representation quality. To achieve efficient visual reinforcement learning, we propose Return-Critic (RC), an auxiliary framework that bridges goal discrepancy by return prediction. RC samples partial frames from an episode, processes them through a shared visual encoder, and employs a lightweight Transformer to predict the episode's return, forcing the encoder to learn return-relevant representation. The attention weights naturally highlight important frames, enabling a key function for prioritized learning. Theoretically, RC can be shown to bridge goal discrepancy, thereby improving representation quality. Extensive experiments on both online (DMControl) and offline (V-D4RL) benchmarks demonstrate that RC significantly enhances the sample efficiency, particularly achieving 68% performance boost on average across nine challenging tasks from DMControl.

Applications · Robotics

Borong Zhang, Jiahao Li, Jiachen Shen, Yishuai Cai, Yuhao Zhang, Yuanpei Chen, Juntao Dai, Jiaming Ji, Yaodong Yang

While Vision-Language-Action models (VLAs) are rapidly advancing toward generalist robot policies, quantitatively characterizing their capability boundaries and failure modes remains challenging. To address this, we introduce **VLA-Arena**, a comprehensive benchmark. It features a novel structured task design framework to quantify difficulty across three orthogonal axes: **(1) Task Structure**, **(2) Language Command**, and **(3) Visual Observation**. This allows us to systematically design tasks with fine-grained difficulty levels, enabling a precise measurement of model capability frontiers. For task structure, VLA-Arena comprises 11 task suites organized into four dimensions: **Safety**, **Distractor**, **Extrapolation**, and **Long Horizon**, totaling 170 tasks. Each suite spans three difficulty levels (L0–L2), with fine-tuning restricted to L0 to rigorously assess generalization. Orthogonal to this, language (W0-W4) and visual (V0-V4) perturbations can be applied to any task as diagnostic probes to distinguish robust grounding from superficial pattern matching. Our extensive evaluation of state-of-the-art VLAs reveals critical limitations: memorization over generalization, superficial visual perception, and a neglect of safety constraints. Additionally, model rank reversals across L0–L2 validate that each level provides non-redundant insights. To foster research addressing these model limitations and ensure reproducibility, we provide the complete VLA-Arena framework, including an end-to-end toolchain from task definition to automated evaluation and the VLA-Arena-S/M/L datasets for fine-tuning. Our benchmark, datasets, models, and leaderboard will be open-sourced.

General Machine Learning · Evaluation

Hanjun Luo, Chiming Ni, Jiaheng Wen, Zhimu Huang, Bingduo Liao, Yiran Wang, Sylvia Chung Yan Shan, Yingbin Jin, Jialin Li, Xinfeng Li 等

LLM-powered coding agents are reshaping the development paradigm. However, existing evaluation systems, neither traditional tests for humans nor benchmarks for LLMs, fail to capture this shift, excluding problems that require both human reasoning to guide solutions and AI efficiency for implementation. We introduce CentaurEval, a unified, ecologically valid benchmark for measuring human-in-the-loop value in coding. CentaurEval's core innovation is its "Collaboration-Necessary" problem templates, which are intractable for standalone LLMs or humans, but solvable through effective collaboration. CentaurEval dynamically instantiates tasks from 45 templates, providing a standardized IDE for humans and a reproducible 450-task toolkit for LLMs. We benchmark 45 participants against 5 LLMs under 4 levels of human intervention. Results show that while LLMs or humans alone achieve poor pass rates (0.67% and 18.89%), human–AI collaboration significantly improves to 31.11%. Our analysis reveals an emerging co-reasoning partnership, challenging the traditional human-tool hierarchy by showing that strategic breakthroughs can originate from either humans or AI. Our work is openly accessible.

General Machine Learning · Transfer, Multitask and Meta-learning

Sander de Haan, Yassine Taoudi-Benchekroun, Pau Vilimelis Aceituno, Benjamin F. Grewe

Catastrophic forgetting remains a fundamental challenge for neural networks when tasks are trained sequentially. In this work, we reformulate continual learning as a control problem where learning and preservation signals compete within neural activity dynamics. We convert regularization penalties into preservation signals that protect prior-task representations. Learning then proceeds by minimizing the control effort required to integrate new tasks while competing with the preservation of prior tasks. At equilibrium, the neural activities produce weight updates that implicitly encode the full prior-task curvature, a property we term the *continual-natural gradient*, requiring no explicit curvature storage. Experiments confirm that our learning framework recovers true prior-task curvature and enables task discrimination, outperforming existing methods on standard benchmarks without replay.