论文检索

Reinforcement Learning · Planning

Yousef Yassin, Junfeng Wen

*AlphaZero* and *MuZero* have demonstrated superhuman performance across a range of strategic tasks. Yet their reliance on maximizing expected returns limits their use in real-world settings, where even high-return policies may incur rare but catastrophic failures. We introduce *RiskZero* to address this limitation; the first *MuZero*-family method for risk-sensitive decision-making, and planning with *zero* prior knowledge of environment dynamics. *RiskZero* learns distributional quantities to estimate trajectory-level risk, guiding search toward policies that explicitly avoid rare but severe outcomes. We establish theoretical convergence to optimal, stationary risk-sensitive policies and validate our approach on environments designed to test risk-sensitive learning from pixels, as well as on larger-scale combinatorial tasks. Across all settings, *RiskZero* consistently outperforms state-of-the-art risk-sensitive baselines, and improves sample efficiency, providing a general framework for safer and reliable model-based reinforcement learning under uncertainty.

General Machine Learning · Methodology

Pankaj Bhagwat, Zhixian Yang, yihao wang, Bei Jiang, Linglong Kong

Conformal Prediction (CP) provides rigorous finite-sample coverage guarantees, yet its statistical efficiency hinges critically on the size of the calibration set. In data-scarce regimes, CP often suffers from volatile quantile estimation, leading to overly conservative and wide prediction intervals. To address this, we propose Random Score Alignment-Conformal Prediction (RSA-CP), a simple framework designed to improve sample efficiency in small-sample CP. Instead of requiring the computationally intensive generation of full synthetic datasets, RSA-CP enhances calibration by directly aligning real scores with a high-resolution reference score distribution. By employing an optimal transport mapping, our framework refines "step-like" quantile increments through a globally optimal use of reference information. We provide theoretical guarantees establishing that RSA-CP maintains robust coverage without any distributional assumptions on the reference scores. Empirical evaluations demonstrate that RSA-CP consistently produces shorter and more precise prediction intervals while maintaining finite-sample coverage guarantees. Overall, RSA-CP offers a computationally efficient and theoretically grounded solution for robust uncertainty quantification under limited data.

Deep Learning · Large Language Models

Haotian Xu, Jiannan Yang, Tian Gao, Lily Weng, Tengfei Ma

Activation sparsity offers a compelling route to accelerate large language model (LLM) inference by selectively suppressing hidden activations, yet existing approaches exhibit severe accuracy degradation at high sparsity. We show that this failure stems from representational instability: *activation sparsity disrupts input-dependent activation learned during pretraining, inducing distribution shifts in hidden states*. We address this issue by reframing activation sparsity as a representational alignment problem and introducing **Spontaneous Neurons (SPON)**, a lightweight mechanism inspired by spontaneous neural activity in biological systems. SPON injects a small set of learnable, input-independent activation vectors that act as persistent representational anchors for sparse computation. These vectors are trained via distribution matching to the dense model and can be absorbed into bias terms after training, incurring negligible inference overhead. Across multiple LLM backbones, SPON consistently restores performance, stabilizes latent representations, and preserves generalization. Our results establish SPON as an effective and principled solution for reliable activation-sparse inference, and offer new insights into knowledge retention in LLMs.

Deep Learning · Large Language Models

Haoyu Zheng, Yongqiang Zhang, Fangcheng Fu, Xiaokai Zhou, Hao Luo, Hongchao Zhu, Yuanyuan Zhu, Hao Wang, Xiao Yan, Jiawei Jiang

To schedule LLM inference, the \textit{shortest job first} (SJF) principle is favorable by prioritizing requests with short output lengths to avoid head-of-line (HOL) blocking. Existing methods usually predict a single output length for each request to facilitate scheduling. We argue that such a \textit{point estimate} does not match the \textit{stochastic} decoding process of LLM inference, where output length is \textit{uncertain} by nature and determined by when the end-of-sequence (EOS) token is sampled. Hence, the output length of each request should be fitted with a distribution rather than a single value. With an in-depth analysis of empirical data and the stochastic decoding process, we observe that output length follows a heavy-tailed distribution and can be fitted with the log-t distribution. On this basis, we propose a simple metric called Tail Inflated Expectation (TIE) to replace the output length in SJF scheduling, which adjusts the expectation of a log-t distribution with its tail probabilities to account for the risk that a request generates long outputs. To evaluate our TIE scheduler, we compare it with three strong baselines, and the results show that TIE reduces the per-token latency by $2.31\times$ for online inference and improves throughput by $1.42\times$ for offline data generation.

Reinforcement Learning · Deep RL

Dohyung Kim, Minbeom Kim, Jeonghye Kim, Lee Sangmook, Sojeong Rhee, Kyomin Jung

Reward-maximizing RL methods enhance the reasoning performance of LLMs, but often reduce the diversity among outputs. Recent works address this issue by adopting GFlowNets, training LLMs to match a target distribution while jointly learning its partition function. In contrast to prior works that treat this partition function solely as a normalizer, we reinterpret it as a per-prompt expected-reward (i.e., online accuracy) signal, leveraging this unused information to improve sample efficiency. Specifically, we first establish a theoretical relationship between the partition function and per-prompt accuracy estimates. Building on this key insight, we propose \textbf{Pa}rtition Fun\textbf{c}tion-Guid\textbf{ed} \textbf{RL} (PACED-RL), a post-training framework that leverages accuracy estimates to prioritize informative question prompts during training, and further improves sample efficiency through an accuracy estimate error–prioritized replay. Crucially, both components reuse information already produced during GFlowNet training, effectively amortizing the compute overhead into the existing optimization process. Extensive experiments across diverse benchmarks demonstrate strong performance improvements over GRPO and prior GFlowNet approaches, highlighting PACED-RL as a promising direction for a more sample efficient distribution-matching training for LLMs.

Deep Learning · Other Representation Learning

Yi Liu, Hongji Zhang, Lei Chen, Mingxuan Yuan, Qiang Xu

Developing effective representations for register transfer level (RTL) designs is crucial for accelerating the hardware design workflow. Existing approaches, however, typically rely on a single data modality, either the RTL code or its associated graph-based representation, limiting the expressiveness and generalization ability of the learned representations. For RTL, the control data flow graph (CDFG) offers a comprehensive structural representation that preserves complete information, while the code modality explicitly encodes semantic and functional information. We argue that integrating these complementary modalities is essential for a thorough understanding of RTL designs. To this end, we propose UniRTL, a multimodal pretraining framework that learns unified RTL representations by jointly leveraging code and CDFG. UniRTL achieves fine-grained alignment between code and graph through mutual masked modeling and employs a hierarchical training strategy that incorporates a pretrained graph-aware tokenizer and staged alignment of text (*i.e.*, functional summary) and code prior to graph integration. We evaluate UniRTL on two downstream tasks, performance prediction and code retrieval, under multiple settings. Experimental results show that UniRTL consistently outperforms prior methods, establishing it as a more robust and powerful foundation for advancing hardware design automation.

Deep Learning · Foundation Models

Yifei He, Yuzheng Hu, Yong LIN, Tong Zhang, Han Zhao

Model merging offers an effective strategy to combine the strengths of multiple finetuned models into a unified model that preserves the specialized capabilities of each. Existing methods merge models in a global manner, performing arithmetic operations across all model parameters. However, such global merging often leads to task interference, degrading the performance of the merged model. In this work, we introduce Localize-and-Stitch, a novel approach that merges models in a localized way. Our algorithm works in two steps: i) Localization: identify tiny ($1\%$ of the total parameters) localized regions in the finetuned models containing essential skills for the downstream tasks, and ii) Stitching: reintegrate only these essential regions back into the pretrained model for task synergy. We demonstrate that our approach effectively locates sparse regions responsible for finetuned performance, and the localized regions could be treated as compact and interpretable representations of the finetuned models (tasks). Empirically, we evaluate our method on various vision and language benchmarks, showing that it outperforms existing model merging methods under different data availability scenarios. Beyond strong empirical performance, our algorithm also facilitates model compression and preserves pretrained knowledge, enabling flexible and continual skill composition from multiple finetuned models with minimal storage and computational overhead.

Social Aspects · Safety

Haozheng Luo, Yimin Wang, Jiahao Yu, Binghui Wang, Yan Chen

We propose CRAFT, a red-teaming alignment framework that leverages model reasoning capabilities and hidden representations to improve robustness against jailbreak attacks. Unlike prior defenses that operate primarily at the output level, CRAFT aligns large reasoning models to generate safety-aware reasoning traces by explicitly optimizing objectives defined over the hidden state space. Methodologically, CRAFT integrates contrastive representation learning with reinforcement learning to separate safe and unsafe reasoning trajectories, yielding a latent-space geometry that supports robust, reasoning-level safety alignment. Theoretically, we show that incorporating latent–textual consistency into GRPO eliminates superficially aligned policies by ruling them out as local optima. Empirically, we evaluate CRAFT on multiple safety benchmarks using two strong reasoning models, Qwen3-4B-Thinking and R1-Distill-Llama-8B, where it consistently outperforms state-of-the-art defenses such as IPO and SafeKey. Notably, CRAFT delivers an average **79.0%** improvement in reasoning safety and **87.7%** improvement in final-response safety over the base models, demonstrating the effectiveness of hidden-space reasoning alignment.

Reinforcement Learning · Deep RL

Haoxi Li, Qinglin Hou, Jianfei Ma, Jinxiang Lai, Tao Han, Sikai Bai, Jingcai Guo, Jie ZHANG, Song Guo

To navigate partially observable visual environments, recent VLM agents increasingly internalize world modeling capabilities directly into their policies via explicit CoT reasoning with reinforcement learning (RL). However, mere passive exploitation of reasoning on visited states is insufficient for sparse-reward agentic tasks, as it lacks the epistemic drive to actively uncover the *known unknown* required for robust generalization. We ask: *Can VLM agents actively find signals that challenge and update their internal world model through curiosity-driven exploration?* In this work, we propose **GLANCE**, a unified framework that bridges reasoning and exploration by grounding the agent's linguistic world model into the stable visual representations of an evolving target network. Crucially, **GLANCE** leverages the discrepancy between linguistic prediction and visual reality as an intrinsic curiosity signal within reinforcement learning, steering the agent to actively explore areas where its internal model is uncertain. Extensive experiments across a series of agentic tasks show the effectiveness of **GLANCE**, and demonstrate that aligning *what the agent thinks* with *what the agent sees* is key to solving complex or sparse agentic tasks.

Social Aspects · Alignment

Utsav Gupta

Machine learning systems increasingly shape attention, work, education, and social life, yet ML research often treats the question "what is this for?" as external, relying on proxies such as accuracy, engagement, or preference satisfaction. This position paper argues that ML research should be guided by explicit, pluralistic models of human purpose, understood as supporting people's capacity to pursue meaningful, self-chosen life projects with agency. The paper proposes three community practices: (i) purpose articulation, a structured "Purpose Statement" that specifies intended beneficiaries, mechanisms, and falsifiable failure modes; (ii) purpose evaluation, which measures impacts on agency and meaning alongside task performance and harm; and (iii) purpose governance, which updates purpose frameworks through transparent, participatory processes to reduce unaccountable value-setting. This framing enables concrete technical research directions, including objective design beyond preference satisfaction, benchmarks for agency and meaning, pluralistic system behavior, and institution-aware alignment. The paper provides stakeholder-differentiated recommendations for researchers, benchmark creators, conference organizers, and funders, and addresses credible objections including value neutrality, feasibility and measurement validity, the claim that harm prevention is sufficient, and risks of ideological capture or paternalism.

Deep Learning · Foundation Models

Kaiyi Luo, Bangjun Wang, Li Zhang, Fanzhang Li, Fei Zhu, Jiaqing Fan

Static fine-tuning paradigms impose rigid structural constraints on foundation models like the Segment Anything Model 2 (SAM2), limiting their adaptability to the varying complexity of specialized downstream tasks. To overcome this limitation, we propose **ACO-MoE-LoRA**, a dynamic framework that introduces an "Evolving-while-Training" strategy by synergizing Ant Colony Optimization (ACO) with a Latent Space Mixture-of-Experts (MoE) architecture. Central to our method is the ACO-ConvLoRA module, which employs a pheromone-guided routing mechanism to actively govern expert selection and topological evolution. By formulating expert assignment as an evolutionary pathfinding problem, this module effectively mitigates the standard routing collapse issue and enables elastic adjustment of LoRA ranks via weight slicing, bridging discrete structural search with continuous parameter training. Extensive experiments across 16 challenging datasets demonstrate that our framework consistently outperforms leading static adapters, while effectively addressing the local optimality limitations of recent dynamic heuristics. This work presents a self-organizing solution that harmonizes swarm intelligence with gradient optimization for efficiently adapting foundation models to specialized domains.

Social Aspects · Safety

Xu Li, Simon Yu, Minzhou Pan, Yiyou Sun, Bo Li, Dawn Song, Xue Lin, Weiyan Shi

LLM-based agents are becoming increasingly capable, yet their safety lags behind. This creates a gap between what agents can do and should do. This gap widens as agents engage in multi-turn interactions and employ diverse tools, introducing new risks overlooked by existing benchmarks. To systematically scale safety testing in multi-turn, tool-realistic settings, we propose a principled taxonomy that transforms single-turn harmful tasks into multi-turn attack sequences. Using this taxonomy, we construct MT-AgentRisk (Multi-Turn Agent Risk Benchmark), the first benchmark to evaluate multi-turn tool-agent safety. Our experiments reveal substantial safety degradation: the Attack Success Rate (ASR) increases by 16.1\% on average across open and closed models in multi-turn settings. To close this gap, we propose ToolShield, a training-free, tool-agnostic, self-exploration defense: when encountering a new tool, the agent autonomously generates test cases, executes them to observe downstream effects, and distills safety experiences for deployment. Experiments show that ToolShield effectively reduces ASR by 30\% on average in multi-turn interactions.

Theory · Learning Theory

Hongxu Chen, Ke Wei, Xiaoming Yuan, Luo Luo

The empirical evidence indicates that stochastic optimization with heavy-tailed gradient noise is more appropriate to characterize the training of machine learning models than that with standard bounded gradient variance noise. Most existing works on this phenomenon focus on the convergence of optimization errors, while the analysis for generalization bounds under the heavy-tailed gradient noise remains limited. In this paper, we develop a general framework for establishing generalization bounds under heavy-tailed noise. Specifically, we introduce a truncation argument to achieve the generalization error bound based on the algorithmic stability under the assumption of bounded $p$th centered moment with $p\in(1,2]$. Building on this framework, we further provide the stability and generalization analysis for several popular stochastic algorithms under heavy-tailed noise, including clipped and normalized stochastic gradient descent, as well as their mini-batch and momentum variants.

Reinforcement Learning · Multi-agent

Kexing Peng, Pengyi Li, tinghuai ma, Jianye Hao

Value factorization eases non-stationarity in MARL, but its static coordination assumptions hinder generalization on long-horizon tasks with shifting dependencies. Prior VQ-VAE methods abstract trajectories yet miss time-varying inter-agent dependencies. We present TACTIC, a CTDE framework with three advances: (i) hierarchical goal decomposition to guide exploration under sparse rewards; (ii) dynamic sparse coordination graphs that adapt dependencies via variance-based TD-error pruning; and (iii) a semantic-conditioned VQ-VAE that discretizes trajectories into coordination classes and maps them to graph-level edge decisions, while also conditioning local policies. A pretrained, frozen goal predictor decouples task recognition from control, preventing gradient interference across coordination abstractions. On SMAC and SUMO, TACTIC delivers state-of-the-art coordination and transfer under sparse rewards and dynamic task structures.

Deep Learning · Large Language Models

Jun Li

As large language models (LLMs) are increasingly composed into heterogeneous multi-agent systems, a fundamental reliability challenge emerges: knowledge and governance **fragment across agents**, leading to composition-dependent behaviors and **linear scaling** of violations. We introduce **Judgment Operators (JO)**, a decision-time framework that adapts corrective projection via precedent memory from agent actions onto admissible sets. JO establishes a *unified projection interface* in which governance constraints $\mathcal{C}$ define the *target admissible set* and corrective precedents $\mathcal{P}$ provide *executable corrective knowledge* for adapting the projection map. The centralized operator $\Pi_J: \mathcal{X} \to \mathcal{X}_J$ implements four-way intervention semantics (*Allow*, *Edit*, *Escalate*, *Deny*), enabling minimal repair without modifying agent internals. We formalize JO as an *adaptive projection operator* and establish guarantees of: (1) **composition-invariant enforcement** with **constant violation probability** (vs. linear scaling without JO); (2) **sublinear mistake accumulation** for online adaptation via JO-A under recurring violations; and (3) **semantic preservation** for code transformation tasks via structure-preserving projection. Empirically, JO provides *portable corrective knowledge transfer*: (1) **capability**---learns and reuses corrective precedents under recurring violations, improving task success over strong baselines; (2) **governance**---achieves *near-perfect constraint enforcement* in fully verifiable settings (0\% observed violation rate vs. 48--68\% for baseline methods); and (3) **portability**---enables *13.5--20.5\% absolute zero-shot cross-model transfer* where few-shot prompting fails. Judgment Operators thus provide a **portable, auditable, and composable interface** for both decision-time governance and capability injection in multi-agent LLM systems, addressing fragmentation at its architectural root through **adaptive, composition-invariant projection**.

Social Aspects · Accountability, Transparency, and Interpretability

Soham Gadgil, Chris Lin, Su-In Lee

Sparse autoencoders (SAEs) are used to decompose neural network activations into human-interpretable features. Typically, features learned by a single SAE are used for downstream applications. However, it has recently been shown that a single SAE captures only a limited subset of features that can be extracted from the activation space. Motivated by this limitation, we introduce and formalize SAE ensembles. Furthermore, we propose to ensemble multiple SAEs through *naive bagging* and *boosting*. In naive bagging, SAEs trained with different weight initializations are ensembled, whereas in boosting SAEs sequentially trained to minimize the residual error are ensembled. Theoretically, naive bagging and boosting are justified as approaches to reduce reconstruction error. Empirically, we evaluate our ensemble approaches with three settings of language models and SAE architectures. Our empirical results demonstrate that, compared to an expanded SAE that matches the number of features in the ensemble, ensembling SAEs improves the reconstruction of language model activations along with SAE stability. Additionally, on downstream tasks such as concept detection and spurious correlation removal, SAE ensembles achieve better performance, showing improved practical utility.

General Machine Learning · Methodology

Hengyi Ren, Yuchen Xie, Changlong Wang, Xin Li, Yue Huang, Jian Guo, Lijuan Sun

Multimodal Federated Learning (MMFL) addresses collaborative training across clients with heterogeneous modality configurations, where effective client selection becomes critical under the compounded challenges of modality, distribution, and quantity heterogeneity. Existing selection methods operate within a reactive paradigm, responding to current observations without anticipating how decisions influence future optimization trajectories. This myopic approach leads to suboptimal convergence when training dynamics shift rapidly under severe heterogeneity. We propose FedSSM, which reconceptualizes client selection as a proactive decision-making process by predicting training dynamics through decision-aware state space models. The prediction error yields a \emph{surprise} signal that quantifies uncertainty and governs adaptive participation budgets and exploration-exploitation trade-offs via counterfactual reasoning over candidate actions. For aggregation, we introduce trust-weighted fusion with modality-specific routing, where surprise calibrates sensitivity to client anomalies. Experiments on four multimodal benchmarks demonstrate that FedSSM achieves 2.5--4.5\% accuracy improvements over state-of-the-art methods while reducing communication rounds by over 30\%.

Applications · Robotics

Zeying Li, shuai zhao, Chaowen Wu, Boyang Li, Kai Huang

Unmanned aerial vehicle (UAV) autopilot systems typically comprise navigation and flight-control modules, and their effective scheduling is critical to achieving high flight performance. However, most existing UAV platforms adopt a split architecture in which navigation and flight control are deployed on separate hardware devices. This separation restricts system-wide observability and prevents holistic scheduling and optimization across the entire autopilot pipeline. Moreover, autonomous flight performance emerges from implicit, cross-coupled, and accumulated interactions among multiple factors, rendering traditional model-based or heuristic scheduling approaches ineffective. To address these challenges, we propose UAV$^2$, a unified and adaptive scheduling framework for UAV autopilot systems with reinforcement learning, targeting flight performance optimization. UAV$^2$ integrates navigation and flight control onto a single onboard computing platform and operating system, formulates the scheduling problem as a partially observable Markov decision process, and learns scheduling policies from runtime execution feedback. The proposed approach is trained and evaluated in a hardware-in-the-loop simulation environment. Experimental results demonstrate that the learned scheduling policy consistently outperforms fixed-rate scheduling strategies in terms of flight robustness and tracking performance.

Reinforcement Learning · Everything Else

Ran Li, Zeyuan Liu, Yinghao Chen, Bingxiang He, Jiarui Yuan, Zixuan Fu, Weize Chen, Jinyi Hu, Chen Qian, Zhiyuan Liu 等

Large Language Models (LLMs) have demonstrated strong potential in complex reasoning, yet their progress remains fundamentally constrained by reliance on massive high-quality human-curated tasks and labels, either through supervised fine-tuning (SFT) or reinforcement learning (RL) on reasoning-specific data. This dependence renders supervision-heavy training paradigms increasingly unsustainable, with signs of diminishing scalability already evident in practice. To overcome this limitation, we introduce CPMöbius, a collaborative Coach–Player paradigm for data-free reinforcement learning of reasoning models. Unlike traditional adversarial self-play frameworks, CPMöbius inspired by multi-agent collaboration treats the Coach and Player as independent but cooperative roles. The Coach proposes instructions targeted at the Player’s capability and receives rewards based on changes in the Player’s performance, while the Player is rewarded for solving the increasingly instructive tasks generated by the Coach. This cooperative optimization loop is designed to directly enhance the Player’s mathematical reasoning ability. Remarkably, CPMöbius achieves substantial improvement without relying on any external training data, outperforming existing unsupervised approaches. For example, on the Qwen2.5-Math-7B-Instruct, our method improves accuracy by overall average +4.9 and out-of-distribution average +5.4, which exceed RENT for +1.5 on overall accuracy and R-zero for +4.2 on OOD accuracy.

Applications · Chemistry, Physics, and Earth Sciences

Mohammad Haddadnia, Yuvan Chali, Abhilash Jayaraj, Constance Kraay, Joana Reis, Felix Strieth-Kalthoff, Haribabu Arthanari

Identifying high-utility candidates from massive discrete spaces under expensive evaluations is a recurring challenge across the sciences, with structure-based drug discovery as a prominent example. While surrogate-based optimization can increase sample efficiency by reducing the number of expensive evaluations, modern molecular libraries have reached billions to trillions of compounds, making full-library surrogate inference itself a major computational bottleneck. We introduce BOBA, a bandit-guided surrogate optimization framework that eliminates full-library inference by adaptively allocating computation across partitions of the action space. By treating partitions as arms in a multi-armed bandit, BOBA concentrates inference and evaluations on empirically promising partitions while maintaining principled exploration. Experiments on real-world synthesis-on-demand libraries demonstrate that optimism-under-uncertainty bandits, combined with meaningful action space partitioning, are essential for effective allocation of inference and evaluations. Our findings reveal a tunable tradeoff between screening performance and surrogate inference cost, which supports practical optimization over current libraries, and establishes a viable route to ultra-large library virtual screening.