论文检索

Deep Learning · Large Language Models

Wenhao Wang, Peizhi Niu, Gongyi Zou, Xiyuan Yang, Jingxing Wang, Haoting Shi, Yaxin Du, Jingyi Chai, Xianghe Pang, shuo tang 等

a transformative standard for connecting large language models (LLMs) with external data sources and tools, and has been rapidly adopted across personal applications and development platforms. However, existing benchmarks predominantly focus on generic information-seeking tools and fail to capture the practical challenges posed by personal social applications, where tools interact with individual accounts or local databases. To bridge this critical gap, we introduce MCP-Persona, the first benchmark specifically designed for evaluating agent performance on real-world, personalized MCP tools. MCP-Persona encompasses a diverse set of widely-used applications, ranging from social media platforms like Reddit and Xiaohongshu (Rednote) to enterprise collaboration suites such as Lark (Feishu) and Slack. Our extensive experiments on various state-of-the-art (SOTA) agents demonstrate their significant struggles with personalized tool use, thereby highlighting the benchmark's crucial role in identifying and addressing these limitations. MCP-Persona is publicly available at \href{https://anonymous.4open.science/r/MCP-Persona-F85D}{https://anonymous.4open.science/r/MCP-Persona-F85D}

General Machine Learning · Evaluation

Marina Mancoridis, Zoe Hitzig

Large language models are often evaluated for correctness on isolated questions. But modern deployments also rely on a different property: whether the model stays consistent as it generates, critiques, and revises over multiple steps that rely on the same underlying concepts. In these settings, *self-consistency* seems desirable, since it limits drift as models reuse and assess their own outputs. Yet we lack standard deployment-relevant ways of quantifying it. How do we measure self-consistency, and what can it signal for reliability in deployment? In this paper we propose a new measure: *generator–evaluator self-consistency*, which assesses whether a model applies the same underlying concept consistently when it is invoked across related prompts. We find that models exhibit substantial variation in self-consistency that is independent of their accuracy on benchmark questions involving those same concepts. Examining this variation in a clinical setting with physician-validated mistakes, we find that higher self-consistency is linked to greater vulnerability to mistakes. Rather than signaling robust understanding, consistency can reflect stable commitment to incomplete or brittle conceptualizations. We interpret this pattern as a *consistency trap*: self-consistency is operationally useful, but can also be evidence of systematized errors.

Deep Learning · Large Language Models

Haichuan Wang, Tao Lin, Lingkai Kong, Ce Li, Hezi Jiang, Milind Tambe

Existing alignment methods directly use the reward model learned from user preference data to optimize an LLM policy, subject to KL regularization with respect to the base policy. This practice is suboptimal for maximizing user's utility because the KL regularization may cause the LLM to inherit the bias in the base policy that conflicts with user preferences. While amplifying rewards for preferred outputs can mitigate this bias, it also increases the risk of reward hacking. This tradeoff motivates the problem of optimally designing reward models under KL regularization. We formalize this reward model optimization problem as a Stackelberg game, and show that a simple reward shaping scheme can effectively approximate the optimal reward model. We empirically evaluate our method in inference-time alignment settings and demonstrate that it integrates seamlessly into existing alignment methods with minimal overhead. Our method consistently improves average reward and achieves win–tie rates exceeding 66\% against all baselines, averaged across evaluation settings.

Deep Learning · Large Language Models

Zexi Liu, Jingyi Chai, Xinyu Zhu, shuo tang, Rui Ye, Weiyu Ma, Bo Zhang, LEI BAI, Siheng Chen

The emergence of large language model (LLM)-based agents has significantly advanced the development of autonomous machine learning (ML) engineering. However, the dominant prompt-based paradigm exhibits limitations: smaller models lack the capacity to learn from execution trajectories for generalization, while large proprietary models incur high computational overhead, restricting accessibility and scalability. Focusing on this, for the first time, we explore the paradigm of learning-based agentic ML, where an LLM agent learns through interactive experimentation on ML tasks using online reinforcement learning (RL). To realize this, we propose a novel agentic ML training framework with three key components: (1) exploration-enriched fine-tuning, which enables LLM agents to generate diverse actions for enhanced RL exploration; (2) step-wise RL, which enables training on a single action step, accelerating experience collection and improving training efficiency; (3) an agentic ML-specific reward module, which unifies varied ML feedback signals into consistent rewards for RL optimization. Leveraging this framework, we train ML-Agent, driven by a 7B-sized Qwen-2.5 LLM for autonomous ML. Despite training on only 9 ML tasks, our 7B-sized ML-Agent achieves comparable performance to agents using much larger proprietary LLMs (e.g., GPT-5) but at significantly lower computational cost, demonstrating strong performance and cross-task generalization.

Deep Learning · Large Language Models

Zhichen Dong, Yang Li, Yuhan Sun, Weixun Wang, Yijia Luo, Zinian Peng, Wenbo Su, YuCheng, Bo Zheng, Junchi Yan

Token-level credit assignment remains a key obstacle for reinforcement learning (RL) in large language models (LLMs), where RL recipes typically treat all tokens equally, failing to distinguish decisive reasoning steps from routine formatting or fluent filler. Recent attempts leverage model-internal signals to assign finer-grained credit, but these are often point-wise heuristics that ignore the global structure of information propagation. We propose FlowTracer, an RL framework that traces \emph{answer-targeted reasoning flow} on an attention-induced directed acyclic graph in which nodes correspond to tokens and edge capacities come from aggregated attention weights and derives token credit from this global structure. The edge capacities are reweighted to retain only the influence that can reach the answer region, while enforcing local flow conservation so intermediate tokens neither lose nor gain effective mass due to path length or irrelevant branches. On this graph, FlowTracer extracts an information-flow backbone connecting the question to the answer and scores tokens by flow throughput, revealing high-impact hubs and aggregation checkpoints that mediate long-range dependencies. These derived importances are used to shape token-level rewards, enabling learning signals to focus precisely on the tokens that route information toward (or away from) correct answers and delivering consistent performance gains across a range of reasoning tasks.

Deep Learning · Large Language Models

Yixuan Xu, Yash Savani, Fei Fang, Zico Kolter

Reinforcement learning with verifiable rewards (RLVR) has emerged as the leading approach for enhancing reasoning capabilities in large language models. However, it faces a fundamental compute and memory asymmetry: rollout generation is embarrassingly parallel and memory-light, whereas policy updates are communication-heavy and memory-intensive. To address this, we introduce PODS (Policy Optimization with Down-Sampling), which decouples rollout generation from policy updates by training only on a strategically selected subset of rollouts, maintaining learning quality while dramatically reducing update costs. We propose a principled subset selection criterion—max-variance down-sampling—that maximizes the variance of reward in the selected subset, and provide an efficient $O(n\log n)$ implementation of this rule. Empirically, Group Relative Policy Optimization (GRPO) coupled with PODS achieves the peak test accuracy of vanilla GRPO at least $\mathbf{1.7\times}$ faster across the different reasoning benchmarks and hardware configurations we tested.

Deep Learning · Robustness

Jingwei Sun, Jianing Zhu, Yuanyi Li, Tongliang Liu, Xia Hu, Bo Han

Autonomous computer use agents that powered by multimodal large language models (MLLMs) are emerging as capable assistants for completing complex digital workflows. However, real-world execution environments are far from ideal: pop-up dialogs, resolution changes, and competing applications frequently interfere with agent perception and control. We introduce AgentHijack, a benchmark designed to evaluate the robustness and reliability of computer-use agents under common corruptions, where the uncertainties in dynamic environment disrupt the execution flow without direct adversarial intent. Specifically, AgentHijack introduces 9 configurable common corruptions to replicate realistic imperfect scenarios. We evaluate a variety of desktop tasks that utilize MLLM-based agents and discover that even minor instances of corruption can result in substantial performance degradation, which emphasizes the fragility of agents and underscores the necessity of robustness evaluation. Afterward, we propose AgentHijack-Agent, a framework that integrates an action generator with enhanced grounding capabilities and an onlooker responsible for behavior summarization and environment checking. Extensive experiments validate its effectiveness. We anticipate that our work will underscore the importance of GUI agents robustness and inspire more follow-up research efforts in this direction.

Social Aspects · Accountability, Transparency, and Interpretability

Biagio La Rosa, Leilani Gilpin

Compositional explanations are a family of methods that aim to describe the spatial alignment between neurons' receptive field activations and concepts through logical rules, typically computed via a search over all possible concept combinations. Since computing the spatial alignment over the entire state space is computationally infeasible, the literature commonly adopts beam search to restrict the space. However, beam search cannot provide any theoretical guarantees of optimality, and it remains unclear how close current explanations are to the true optimum. In this theoretical paper, we address this gap by introducing the first framework for computing guaranteed optimal compositional explanations. Specifically, we propose: (i) a decomposition that identifies the factors influencing the spatial alignment, (ii) a heuristic to estimate the alignment at any stage of the search, and (iii) the first algorithm that can compute optimal compositional explanations within a feasible time. Using this framework, we demonstrate that 10-40% of explanations previously obtained with beam search are suboptimal when overlapping concepts are involved. Finally, we evaluate a beam-search variant guided by our proposed decomposition and heuristic, showing that it matches or improves runtime over prior methods while offering greater flexibility in hyperparameters and computational resources.

Social Aspects · Privacy

Ahmed Mehdi Inane, Vincent Quirion, Gintare Karolina Dziugaite, Ioannis Mitliagkas

Noise-based certified machine unlearning currently faces a hard ceiling: the noise magnitude required to certify unlearning typically destroys model utility, particularly for large-scale deletion requests. While leveraging public data is a standard technique in differential privacy to relax this tension, its role in unlearning remains unexplored. We address this gap by introducing **Asymmetric Langevin Unlearning (ALU)**, a framework that uses public data to mitigate privacy costs. We prove that public data injection suppresses the unlearning cost by a factor of $O(1/n_{\mathrm{pub}}^2)$, guaranteeing a strict computational advantage over retraining. This establishes a new control mechanism: practitioners can mitigate the need for high noise—and the associated utility loss—by increasing the volume of public data. Crucially, we analyze the realistic setting of **distribution mismatch**, explicitly characterizing how shifts between public and private sources impact utility. We show that ALU enables "mass unlearning'' of constant dataset fractions -- a regime where standard symmetric methods become impractical -- while maintaining high utility. Empirical evaluations using variational Rényi divergence and membership inference attacks confirm that ALU effectively thwarts privacy attacks while preserving utility under reasonable distribution shifts.

Deep Learning · Large Language Models

Qingyao Li, Weiwen Liu, Weinan Zhang, Yong Yu, Bo An

Recent advancements in Large Language Models (LLMs) have successfully employed search-based strategies to enhance code generation. However, existing methods typically rely on static, sparse public test cases for verification, leading to pseudo-correctness—where solutions overfit the visible public tests but fail to generalize to hidden test cases. We argue that optimizing against a fixed, weak environment inherently limits robustness. To address this, we propose AdverMCTS, a novel adversarial Monte Carlo Tree Search framework that combats pseudo-correctness by coupling code search with active vulnerability discovery. AdverMCTS formulates generation as a minimax-style game between a Solver agent, which synthesizes code candidates, and an Attacker agent, which evolves to generate targeted test cases that exploit logical divergences in the current solution pool. These discovered tests form a dynamic, progressively hostile filter that penalizes fragile reasoning. Extensive experiments demonstrate that AdverMCTS significantly outperforms state-of-the-art baselines, effectively reducing false positive rates and forcing the model to generalize beyond the initial constraints. The resources of this work are available at https://anonymous.4open.science/r/AdverMCTS_open-A255.

Deep Learning · Generative Models and Autoencoders

Dong Hoon Lee, Seunghoon Hong

Latent Diffusion Models (LDMs) have become dominant in visual synthesis, but their quality–compute trade-off is largely constrained by the tokenizer’s fixed compression ratio. Variable-length tokenizers (VLTs) promise adaptive compression by varying token counts, allowing diffusion models to flexibly balance quality and compute. However, conventional VLTs modulate length by truncating ordered token sequences, which changes token semantics across lengths and breaks representational alignment. This leads to significant cross-length variation in the latent distribution that hinders a single variable-length diffusion model from operating effectively. To address this, we propose a novel variable-length tokenizer that modulates length by merging tokens. We show that encouraging similar tokens to merge enables direct cross-length representation alignment when the diffusion transformer operates according to the merging pattern. Since conventional merging methods are data-dependent, making the merging pattern inaccessible during generation, we introduce learnable global merging, which is data-independent, to ensure compatibility with diffusion transformers. On ImageNet 256$\times$256 generation, our merging-based variable-length tokenizer integrated with a diffusion transformer achieves a superior gFID–compute trade-off compared to prior VLT methods.

General Machine Learning · Causality

Kenneth Lee, Zihan Zhou, Murat Kocaoglu

Modern cloud systems rely on architectures with many interconnected microservices, which enable scalability and flexibility but make troubleshooting failures difficult. Identifying the root cause requires navigating complex dependencies, often beyond the capacity of domain experts. Causal models offer a principled approach to root cause analysis (RCA), but prior methods are typically sample inefficient, as they assume access to the full causal graph or require large numbers of post-failure interventions. We introduce Bayesian Root Cause Discovery (BRCD), which leverages a partial causal structure (a CPDAG learned during the pre-failure period) and performs Bayesian inference without enumerating all DAGs from each interventional Markov equivalence class ($\mathcal{I}$-MEC) for each root cause candidate. Using a recent uniform DAG sampling framework (Wienöbst et al., 2023), BRCD provides the first statistical consistency guarantees for nonparametric RCA, with both identifiability and finite-sample posterior bounds under $\varepsilon$-vanishing approximation. Empirically, across synthetic benchmarks and three microservice systems (Online Boutique, Sockshop, Petshop), BRCD achieves state-of-the-art top-$l$ accuracy while remaining effective in low-failure-sample regimes and scaling to large graphs.

Deep Learning · Theory

Yang Yang, Zhengmin Kong, Yuan Liu, Tao Huang, Wei Xiang

Modern neural networks derive much of their effectiveness from rich connectivity patterns. Yet, existing architectures often fix the topology at either the sparse or dense extremes, thereby limiting structural flexibility and analysis. We propose Kronecker Generative Networks (KGNs), an algebraic framework that constructs neural network topologies via recursive generation rules, treating topology as a first-class design object. KGNs generate families of directed acyclic graphs with controllable connectivity complexity, enabling systematic interpolation between sparse and dense aggregation regimes. Under this formulation, architectures such as FractalNet and DenseNet arise as specific instantiations corresponding to different generation rules. We provide theoretical analysis of acyclicity, connectivity scaling, and expressiveness, and demonstrate experimentally that KGN instantiations achieve favorable accuracy-efficiency trade-offs across multiple domains.

General Machine Learning · Clustering

Fan Yang, Haikun Xu

Incomplete multi-view clustering (IMVC) becomes particularly challenging under heavy missingness and view imbalance, where scarce co-observed pairs make cross-view correspondences unreliable: imputation-first pipelines can trigger cascading reconstruction errors, while purely consistency-based alignment often degrades sharply and offers limited control over semantic convergence across views. We propose \textbf{MAGIC} (Masked multi-p\textbf{A}th contrast with conf\textbf{I}dence-\textbf{G}ated semant\textbf{I}c imputation), a unified framework that learns calibrated cluster semantics before performing conservative completion. MAGIC instantiates multiple correlated representation and prediction paths from lightly augmented latent codes and couples them via a masked multi-path contrastive consensus objective with prediction-consistency regularization, yielding stable posteriors even when co-observations are scarce; these posteriors are then aggregated into view-wise soft assignments to reduce overconfidence and alleviate dominance by highly available views. Building on the calibrated semantics, MAGIC conducts similarity-guided semantic transfer in label space with confidence-aware gating and completes missing representations in a geometry-preserving manner, thereby mitigating error propagation under severe missingness. Extensive experiments on four benchmarks across a wide range of missing ratios demonstrate consistent improvements over prior IMVC methods, and ablations validate the complementary roles of masked multi-path consensus learning and confidence-gated semantic imputation.

Social Aspects · Fairness

David Troxell, Noah Roemer, Guido Montufar

Differentiable optimization layers are traditionally integrated in predict-then-optimize frameworks where a neural model estimates parameters that subsequently serve as fixed inputs to downstream decision-making optimization problems. In this work, we introduce the concept of a ``fairness layer'': a differentiable optimization layer appended to a model's output layer that guarantees a chosen notion of output parity is satisfied when integrated into a neural network. Additionally, we introduce an online primal-dual inference algorithm that provides provable aggregate fairness guarantees for streaming predictions with arbitrarily small batch sizes, where traditional per-batch constraints become overly restrictive. Numerical experiments demonstrate the effectiveness of the fairness layer and associated algorithm, and theoretical analysis characterizes the layer's differentiability and stability properties during model training and backpropagation. Our code for these experiments is publicly available on GitHub: https://github.com/anonymouspapersubmission012345/icml_2026_submission and our public Python package documentation can be found online: https://anonymouspapersubmission012345.github.io/fairness_training_anonymous/ .

Probabilistic Methods · Gaussian Processes

Donghyun Lee, Young Myoung Ko

Modeling the learning curve is critical for cost-effective data collection in deep learning systems. Most prior approaches assume a specific parametric learning curve, but these can be inappropriate when no reliable parametric form can be assumed for the learning curve. While Gaussian processes offer flexible nonparametric modeling, existing GP approaches that enforce monotonicity typically introduce intractable factors or require derivative observations. To address this, we propose a Monotonic Variational Gaussian Process for Efficient Data Collection (MOVE), which (i) introduces a novel monotonic variational GP formulation with virtual-derivative factors to enable tractable posterior inference, and (ii) develops an expected shortfall based objective for target-driven data collection. Furthermore, our theoretical analysis shows that expected shortfall provides non-vanishing gradient signals that enable reliable gradient-based optimization. Extensive experiments on classification, segmentation, and detection benchmarks demonstrate consistent improvements over the prior method.

Deep Learning · Generative Models and Autoencoders

Jiawei Zhang, Ziyuan Liu, Leon Yan, Zhenyu Xiao, Yuantao Gu

The distortion–perception (D–P) tradeoff is a fundamental phenomenon of Bayesian inverse problems, which characterizes the inherent tension between distortion performance and perceptual quality. Enabling flexible traversal of the D-P tradeoff at inference time is crucial for practical applications. Despite the recent success of diffusion models in zero-shot inverse problem solving, efficient and principled strategies for D-P traversal in diffusion-based inverse algorithms remain inadequately characterized. In this paper, we propose a stage-wise framework for realizing D-P traversal using a single diffusion model in zero-shot inverse problems. Our proposed method, termed MAP-RPS, starts with an MAP estimation stage that approximates the MMSE solution and provides a low-distortion initialization, followed by a re-noised posterior sampling stage that progressively improves perceptual quality. We provide theoretical analyses for both stages, establishing the validity and effectiveness of the proposed design. Furthermore, we extend MAP-RPS to the latent space, yielding LMAP-RPS, which enjoys broader applicability by leveraging large-scale pre-trained latent diffusion backbones. Extensive experiments demonstrate that MAP-RPS and LMAP-RPS enable more effective D-P traversal on various tasks, while also exhibiting strong performance as efficient solvers for real-world inverse problems.

Applications · Everything Else

Aurelien Ghiglino, Daniel Elenius, Anirban Roy, Ramneet Kaur, Manoj Acharya, Colin Samplawski, Brian Matejek, Susmit Jha, Juan Alonso, Adam Cobb

In this paper, we generate conceptual engineering designs of electric vertical take-off and landing (eVTOL) aircraft. We follow the paradigm of simulation-based inference (SBI), whereby we look to learn a posterior distribution over the full eVTOL design space. To learn this distribution, we must sample over discrete aircraft configurations (topologies) and their corresponding set of continuous parameters. Therefore, we introduce a hierarchical probabilistic model consisting of two diffusion models. The first model leverages recent work on Riemannian Diffusion Language Modeling (RDLM) and Unified World Models (UWMs) to enable us to sample topologies from a discrete and continuous space. For the second model we introduce a masked diffusion approach to sample the corresponding parameters conditioned on the topology. We show our approach successfully rediscovers known trends and captures governing physical laws in aircraft design.

Reinforcement Learning · Deep RL

Giorgi Butbaia, Paul Orland, Coco Huang, Davide Passaro, Lucas Fagan, Michele Tarquini, Hailong Dao, David Eisenbud, Ali Shehper, Sergei Gukov

Applying machine learning techniques to solving long-standing mathematical conjectures can be particularly challenging due to their extreme reward sparsity. As an illustrative example, we consider Kalai's algebraic Hirsch conjecture and recast the construction of its counterexamples as a sparse-reward reinforcement learning problem on graphs. We propose a constrained options-based HRL framework with an equivariant graph neural network policy, which allows us to learn useful temporal abstractions for this task. We evaluate our approach over a wide range of degrees and demonstrate that it consistently outperforms classical RL algorithms as well as greedy search. By exploiting the hierarchical structure of the problem, we effectively provide a first-of-its-kind application of HRL to a problem in commutative algebra.

Optimization · Large Scale, Parallel and Distributed

Nozomi Hata, Kenta Niwa

For communication-efficient decentralized learning, advanced network (NW) topologies, such as exponential and 1-peer exponential graphs, have been studied under homogeneous communication delays. However, real-world NWs exhibit heterogeneous communication delays, making node assignment optimization crucial for minimizing the Bottleneck Communication Delay (BCD). We propose BTSP-MSR, an approximate method for minimizing BCD on circulant digraphs, including exponential and 1-peer exponential graphs. Leveraging the fact that circulant digraphs can be viewed as a union of (directed) ring graphs, we derive an upper bound on the BCD by combining the ring-graph BCD (BTSP) with a deviation term (MSR). We then construct a solver that sequentially minimizes these two terms. Numerical experiments show that BTSP-MSR consistently reduces BCD across several circulant digraphs with large numbers of nodes. Notably, incorporating the exponential or 1-peer exponential graph enables communication-efficient decentralized learning under heterogeneous delay settings.