ICLR 2026 Papers

While Language Models (LMs) have made significant progress in automating machine learning engineering (MLE), the acquisition of high-quality MLE training data is significantly constrained. Current MLE benchmarks suffer from low scalability and limited applicability because they rely on static, manually curated tasks that demand extensive time and manual effort to produce. We introduce MLE-Smith, a fully automated multi-agent pipeline, to transform raw datasets into competition-style MLE challenges through an efficient generate--verify--execute paradigm for scaling MLE tasks with verifiable quality, real-world usability and rich diversity. The proposed multi-agent pipeline in MLE-Smith drives structured task design and standardized refactoring, coupled with a hybrid verification mechanism that enforces strict structural rules and high-level semantic soundness. It further validates empirical solvability and real-world fidelity through interactive execution. We apply MLE-Smith to 224 of real-world datasets and generates 606 tasks spanning multiple categories, objectives, and modalities, demonstrating that MLE-Smith can work effectively across a wide range of real-world datasets. Evaluation on generated tasks shows that the performance of eight mainstream and cutting-edge LLMs on MLE-Smith tasks is strongly correlated with their performance on carefully human-designed tasks, highlighting the effectiveness of the MLE-Smith in scaling up MLE tasks while maintaining task quality.

While Large Language Models (LLMs) have shown remarkable advancements in reasoning and tool use, they often fail to generate optimal, grounded solutions under complex constraints. Real-world travel planning exemplifies these challenges, evaluating agents’ abilities to handle constraints that are explicit, implicit, and even evolving based on interactions with dynamic environments and user needs. In this paper, we present ATLAS, a general multi-agent framework designed to effectively handle such complex nature of constraints awareness in real-world travel planning tasks. ATLAS introduces a principled approach to address the fundamental challenges of constraint-aware planning through dedicated mechanisms for dynamic constraint management, iterative plan critique, and adaptive interleaved search. ATLAS demonstrates state-of-the-art performance on the TravelPlanner benchmark, improving the final pass rate from 23.3% to 44.4% over its best alternative. More importantly, our work is the first to demonstrate quantitative effectiveness on real-world travel planning tasks with live information search and multi-turn feedback. In this realistic setting, ATLAS showcases its superior overall planning performance, achieving an 84% final pass rate which significantly outperforms baselines including ReAct (59%) and a monolithic agent (27%).

Despite multi-view clustering has achieved great progress in past decades, it is still a challenge to balance the effectiveness, efficiency, completeness and consistency of nonlinearly separable clustering for the data from different views. To address this challenge, we show that multi-view clustering can be regarded as a three-level optimization problem. To be specific, we divide the multi-view clustering into three sub-problems based on $ K $-means or kernel $ K $-means, i.e., linear clustering on the original multi-view dataset, nonlinear clustering on the set of obtained linear clusters and multi-view clustering by integrating partition matrices from different views obtained by linear and nonlinear clustering based on reconstruction. We propose Explainable $ K $-means Neural Networks (EKNN) and present how to unify these three sub-problems into a framework based on EKNN. It is able to simultaneously consider the effectiveness, efficiency, completeness and consistency for the nonlinearly multi-view clustering and can be optimized by an iterative algorithm. EKNN is explainable since the effect of each layer is known. To the best of our knowledge, this is the first attempt to balance the effectiveness, efficiency, completeness and consistency by dividing the multi-view clustering into three different sub-problems. Extensive experimental results demonstrate the effectiveness and efficiency of EKNN compared with other methods for multi-view clustering on different datasets in terms of different metrics.

Sequential activation of place-tuned neurons in an animal during navigation is typically interpreted as reflecting the sequence of input from adjacent positions along the trajectory. More recent theories about such place cells suggest sequences to arise from abstract cognitive objectives like planning. Here, we propose a mechanistic and parsimonious interpretation to complement these ideas: hippocampal sequences arise from intrinsic recurrent circuitry that propagates activity without sustained input, acting as a temporal memory buffer for extremely sparse inputs. We implement a minimal sequence generator inspired by neurobiology and pair it with an actor–critic learner for egocentric visual navigation. Our agent reliably solves a continuous maze without explicit geometric cues, with performance depending on the length of the recurrent sequence. Crucially, the model outperforms LSTM cores under sparse input conditions (16 channels, $\sim2.5$% activity), but not under dense input, revealing a strong interaction between representational sparsity and memory architecture. Hidden units develop localized place fields, distance-dependent spatial kernels, and task-dependent remapping, while inputs to the generator orthogonalize and spatial information increases across layers. These phenomena align with neurobiological data and are causal to performance. Together, our results show that sparse input synergizes with sequence-generating dynamics, providing both a mechanistic account of place cell sequences in the mammalian hippocampus and a simple inductive bias for reinforcement learning based on sparse egocentric inputs in navigation tasks.

Autonomous agents that operate computers via Graphical User Interfaces (GUIs) often struggle with efficiency and reliability on complex, long-horizon tasks. While augmenting these agents with planners can improve task decomposition, they remain constrained by the inherent limitations of performing all actions through GUI manipulation, leading to brittleness and inefficiency. In this work, we introduce a more robust and flexible paradigm: enabling agents to use coding as an enhanced action. We present CoAct-1, a novel multi-agent system that synergistically combines GUI-based control with direct programmatic execution. CoAct-1 features an Orchestrator that dynamically delegates subtasks to either a conventional GUI Operator or a specialized Programmer agent, which can write and execute Python or Bash scripts. This hybrid approach allows the agent to bypass inefficient GUI action sequences for tasks like file management and data processing, while still utilizing visual interaction when necessary. We evaluate our system on the challenging OSWorld and WindowsAgentArena benchmark, where CoAct-1 achieves a new state-of-the-art success rate of 60.8% on OSWorld and 52.5% on WindowsAgentArena, significantly outperforming prior methods. Furthermore, our approach dramatically improves efficiency, reducing the average number of steps required to complete a task to just 10.15 on OSWorld, compared to 15 for leading GUI agents. Our results demonstrate that integrating coding as a core action provides a more powerful, efficient, and scalable path toward generalized computer automation.

Despite improvements by length extrapolation, efficient attention and memory modules, handling infinitely long documents without performance degradation during extrapolation remains the ultimate challenge in long-text processing. To solve this problem, We introduce a novel agent workflow, \method, which processes text in segments and updates memory through an overwrite strategy, addressing the challenge of long-context task through enhanced memory management. We further extend the DAPO algorithm to directly optimize memory ability in an end-to-end fashion, facilitating training via independent-context multi-conversation generation. Experimental results demonstrate that MemAgent has superb long-context capabilities, being able to extrapolate from an 8K context to a 3.5M QA task with a performance loss of less than 10\% and achieving over 95\% on the 512K NIAH test.

Self-evolution is a central research topic in enabling large language model (LLM)-based agents to continually improve their capabilities after pretraining. Recent research has witnessed a transition from reinforcement learning (RL)-free to RL-based methods. Current RL-based methods either rely on dense external reward signals or extract intrinsic reward signals from LLMs themselves. However, these approaches diverge from the self-evolution mechanisms observed in human intelligence, where individuals learn and improve through mutual discussion and collaboration. In this work, we introduce Co-Evolving Multi-Agent Systems (CoMAS), a novel framework that enables agents to improve autonomously by learning from inter-agent interactions without external supervision. CoMAS generates intrinsic rewards from rich discussion dynamics, employs an LLM-as-a-judge mechanism to formulate these rewards, and optimizes each agent's policy through RL, thereby enabling decentralized and scalable co-evolution. Experimental results demonstrate that CoMAS consistently outperforms untrained agents and achieves state-of-the-art performance across most evaluation settings. Ablation studies confirm the necessity of interaction-based reward signals and reveal promising scalability as the number and diversity of agents increase. These findings establish CoMAS as a novel and effective paradigm for self-evolution in LLM-based agents.

Information Bottleneck (IB) Theory is renowned for its ability to learn simple, compact, and effective data representations. In multi-modal clustering, IB theory effectively eliminates interfering redundancy and noise from multi-modal data, while maximally preserving the discriminative information. Existing IB-based multi-modal clustering methods suffer from low-quality pseudo-labels and over-reliance on accurate Mutual Information (MI) estimation, which is known to be challenging. Moreover, unreliable or noisy pseudo-labels may lead to an overconfident clustering outcome. To address these challenges, this paper proposes a novel CaLibrated Information Bottleneck (CLIB) framework designed to learn a clustering that is both accurate and trustworthy. We build a parallel multi-head network architecture—incorporating one primary cluster head and several modality-specific calibration heads—which achieves three key goals: namely, calibrating for the distortions introduced by biased MI estimation thus improving the stability of IB, constructing reliable target variables for IB from multiple modalities and producing a trustworthy clustering result. Notably, we design a dynamic pseudo-label selection strategy based on information redundancy theory to extract high-quality pseudo-labels, thereby enhancing training stability. Experimental results demonstrate that our model not only achieves competitive clustering accuracy on multiple benchmark datasets but also exhibits excellent performance on the expected calibration error metric. Code is available at \textcolor{red}{https://shizhehu.github.io/}.

Large Language Models (LLMs) have recently been widely adopted in conversational agents. However, the increasingly long interactions between users and agents accumulate extensive dialogue records, making it difficult for LLMs with limited context windows to maintain a coherent long-term dialogue memory and deliver personalized responses. While retrieval-augmented memory systems have emerged to address this issue, existing methods often depend on single-granularity memory segmentation and retrieval. This approach falls short in capturing deep memory connections, leading to partial retrieval of useful information or substantial noise, resulting in suboptimal performance. To tackle these limits, we propose MemGAS, a framework that enhances memory consolidation by constructing multi-granularity association, adaptive selection, and retrieval. MemGAS is based on multi-granularity memory units and employs Gaussian Mixture Models to cluster and associate new memories with historical ones. An entropy-based router adaptively selects optimal granularity by evaluating query relevance distributions and balancing information completeness and noise. Retrieved memories are further refined via LLM-based filtering. Experiments on four long-term memory benchmarks demonstrate that MemGAS outperforms state-of-the-art methods on both question answer and retrieval tasks, achieving superior performance across different query types and top-K settings. \footnote{https://anonymous.4open.science/r/MemGAS-626C/}

We introduce Robust Multi-Objective Decoding (RMOD), a novel inference-time algorithm that robustly aligns Large Language Models (LLMs) to multiple human objectives (e.g., instruction-following, helpfulness, safety) by maximizing the worst-case rewards. RMOD formulates the robust decoding problem as a maximin two-player game between adversarially computed reward weights and the sampling policy, solvable through a Nash equilibrium. We demonstrate that this game reduces to a convex optimization problem to identify the worst-case reward weights, with the optimal sampling policy analytically derived. For practical applications, we propose an efficient algorithm of RMOD tailored for contemporary LLMs, introducing minimal computational overhead compared to standard non-robust Controlled Decoding methods. Experimental results across a range of popular alignment datasets with up to 10 objectives show the effectiveness of RMOD and its distilled version, consistently outperforming baselines in worst-case rewards and win rates.

Deep research---producing comprehensive, citation-backed reports by searching across hundreds of live websites---marks an important frontier for agentic systems. To rigorously evaluate this ability, three principles are essential: tasks should be (1) user-centric, reflecting realistic information needs, (2) dynamic, requiring up-to-date information beyond parametric knowledge, and (3) unambiguous, ensuring consistent interpretation across users. Existing benchmarks fall short of these principles, often focusing on narrow domains or posing ambiguous questions that hinder fair comparison. Guided by these principles, we introduce LiveResearchBench, a benchmark of 100 expert-curated tasks spanning daily life, enterprise, and academia, each requiring extensive, dynamic, real-time web search and synthesis. Built with over 1,500 hours of human labor, LiveResearchBench provides a rigorous basis for systematic evaluation. To evaluate citation-grounded long-form reports, we present DeepEval, a comprehensive suite covering both content- and report-level quality: checklists for coverage and presentation, rubric-tree assessments of citation accuracy and traceability, and metrics for consistency and depth of analysis. Using LiveResearchBench and DeepEval, we conduct a comprehensive evaluation of frontier deep research systems, including single-agent web search, single-agent deep research, and multi-agent systems. Our analysis reveals current strengths, recurring failure modes, and key system components needed to advance reliable, insightful deep research.

In clinical practice, physicians refrain from making decisions when patient information is insufficient. This behavior, known as abstention, is a critical safety mechanism preventing potentially harmful misdiagnoses. Recent investigations have reported the application of large language models (LLMs) in medical scenarios. However, existing LLMs struggle with the abstentions, frequently providing overconfident responses despite incomplete information. This limitation stems from conventional abstention methods relying solely on model self-assessments, which lack systematic strategies to identify knowledge boundaries with external medical evidences. To address this, we propose \textbf{KnowGuard}, a novel \textit{investigate-before-abstain} paradigm that integrates systematic knowledge graph exploration for clinical decision-making. Our approach consists of two key stages operating on a shared contextualized evidence pool: 1) an evidence discovery stage that systematically explores the medical knowledge space through graph expansion and direct retrieval, and 2) an evidence evaluation stage that ranks evidence using multiple factors to adapt exploration based on patient context and conversation history. This two-stage approach enables systematic knowledge graph exploration, allowing models to trace structured reasoning paths and recognize insufficient medical evidence. We evaluate our abstention approach using open-ended multi-round clinical benchmarks that mimic realistic diagnostic scenarios, assessing abstention quality through accuracy-efficiency trade-offs beyond existing closed-form evaluations. Experimental evidence clearly demonstrates that KnowGuard outperforms state-of-the-art abstention approaches, improving diagnostic accuracy by 3.93\% through effective diagnostic interactions averaging 5.74 conversation turns.

Large language models (LLMs) are increasingly deployed in multi-agent systems (MAS) as components of collaborative intelligence, where peer interactions dynamically shape individual decision-making. While prior work has largely focused on conformity bias, we broaden the scope to examine how LLMs build rapport from previous interactions, resist misinformation, and integrate peer input during collaboration, which are key factors for achieving collective intelligence under complex social dynamics. We introduce KAIROS, a benchmark simulating quiz contests with peer agents of varying reliability, offering fine-grained control over conditions such as expert–novice roles, noisy crowds, and adversarial peers. LLMs receive both historical interactions and current peer responses, allowing systematic investigation into how rapport, peer action, and self-confidence influence decisions. To mitigate this vulnerability, we evaluate prompting, supervised fine-tuning, and reinforcement learning using Group Relative Policy Optimization (GRPO) across multiple models. Our results show that model size plays a central role in moderating susceptibility to social influence: larger models exhibit stronger resilience and benefit from prompting-based mitigation, whereas smaller models are more vulnerable. For the latter, carefully configured GRPO training improves both robustness and overall performance. Our code and datasets are available at: https://anonymous.4open.science/r/KAIROS-4F71

In autonomous driving, multi-agent collaborative perception enhances sensing capabilities by enabling agents to share perceptual data. A key challenge lies in handling heterogeneous features from agents equipped with different sensing modalities or model architectures, which complicates data fusion. Existing approaches often require retraining encoders or designing interpreter modules for pairwise feature alignment, but these solutions are not scalable in practice. To address this, we propose GT-Space, a flexible and scalable collaborative perception framework for heterogeneous agents. GT-Space constructs a common feature space from ground-truth labels, providing a unified reference for feature alignment. With this shared space, agents only need a single adapter module to project their features, eliminating the need for pairwise interactions with other agents. Furthermore, we design a fusion network trained with contrastive losses across diverse modality combinations. Extensive experiments on simulation datasets (OPV2V and V2XSet) and a real-world dataset (RCooper) demonstrate that GT-Space consistently outperforms baselines in detection accuracy while delivering robust performance. Our code will be released at https://github.com/KingScar/GT-Space.

Recent vision-language models have strong perceptual ability but their implicit reasoning is hard to explain and easily generates hallucinations on complex queries. Compositional methods improve interpretability, but most rely on a single agent or hand-crafted pipeline and cannot decide when to collaborate across complementary agents or compete among overlapping ones. We introduce MATA (Multi-Agent hierarchical Trainable Automaton), a multi-agent system presented as hierarchical finite-state automaton for visual reasoning whose top-level transitions are chosen by a trainable hyper agent. Each agent corresponds to a state in the hyper automaton, and runs a small rule-based sub-automaton for reliable micro-control. All agents read and write a shared memory, yielding transparent execution history. To supervise the hyper agent’s transition policy, we build transition-trajectory trees and transform to memory-to-next-state pairs, forming the MATA-SFT-90K dataset for supervised finetuning (SFT). The finetuned LLM as the transition policy understands the query and the capacity of agents, and it can efficiently choose the optimal agent to solve the task. Across multiple visual reasoning benchmarks, MATA achieves the state-of-the-art results compared with monolithic and compositional baselines. The code and dataset are available at https://github.com/ControlNet/MATA.

Neural network certification methods heavily rely on convex relaxations to provide robustness guarantees. However, these relaxations are often imprecise: even the most accurate single-neuron relaxation is incomplete for general ReLU networks, a limitation known as the *single-neuron convex barrier*. While multi-neuron relaxations have been heuristically applied to address this issue, two central questions arise: (i) whether they overcome the convex barrier, and if not, (ii) whether they offer theoretical capabilities beyond those of single-neuron relaxations. In this work, we present the first rigorous analysis of the expressiveness of multi-neuron relaxations. Perhaps surprisingly, we show that they are inherently incomplete, even when allocated sufficient resources to capture finitely many neurons and layers optimally. This result extends the single-neuron barrier to a *universal convex barrier* for neural network certification. On the positive side, we show that completeness can be achieved by either (i) augmenting the network with a polynomial number of carefully designed ReLU neurons or (ii) partitioning the input domain into convex sub-polytopes, thereby distinguishing multi-neuron relaxations from single-neuron ones which are unable to realize the former and have worse partition complexity for the latter. Our findings establish a foundation for multi-neuron relaxations and point to new directions for certified robustness, including training methods tailored to multi-neuron relaxations and verification methods with multi-neuron relaxations as the main subroutine.

Attention is foundational to large language models (LLMs), enabling different heads to have diverse focus on relevant input tokens. However, learned behaviors like attention sinks, where the first token receives the most attention despite limited semantic importance, suggest some heads may be inactive, and point to a significant source of computational redundancy. To analyze this phenomenon, we propose a taxonomy of 12 score functions that measure different ways a head can be inactive. Thresholding these scores allows us to analyze different sets of potentially inactive attention heads. We evaluate whether identified heads are inactive through model interventions, finding that more than 12% of attention heads are inactive on average, and can be ablated in specific contexts while maintaining MMLU accuracy to within 1% of the pretrained LLM. Across 3 model families, our score functions that measure the average norm of a head's output consistently identify inactive heads that would not have been found by score functions that rely solely on attention weights. We establish that relying on a score function that measures a first token attention sink would underestimate the prevalence of inactive heads, failing to identify more than 7% of inactive heads on average. We also show how measuring score distributions can provide insights into attention behavior. For instance, we find evidence that finetuning causes little to no change in attention behavior, and that even within the same model family, large model scales present markedly different attention behaviors.

We investigate the challenge of applying Large Language Models (LLMs) to long texts. We propose a theoretical framework that distinguishes the failure modes of long context tasks into three categories: cross-chunk dependence (task noise), confusion that grows with context size (model noise), and the imperfect integration of partial results (aggregator noise). Under this view, we analyze when it is effective to use multi-agent chunking, i.e., dividing a length sequence into smaller chunks and aggregating the processed results of each chunk. Our experiments on tasks such as retrieval, question answering, and summarization confirm both the theoretical analysis and the conditions that favor multi-agent chunking. By exploring superlinear model noise growth with input length, we also explain why, for large inputs, a weaker model configured with chunk-based processing can surpass a more advanced model like GPT4o applied in a single shot. Overall, we present a principled understanding framework and our results highlight a direct pathway to handling long contexts in LLMs with carefully managed chunking and aggregator strategies.

Enzyme Commission (EC) number prediction is vital for elucidating enzyme functions and advancing biotechnology applications. However, current methods struggle to capture the hierarchical relationships among enzymes and often overlook critical structural and active site features. To bridge this gap, we introduce PoinnCARE, a novel framework that jointly encodes and aligns multi-modal data from enzyme sequences, structures, and active sites in hyperbolic space. By integrating graph diffusion and alignment techniques, PoinnCARE mitigates data sparsity and enriches functional representations, while hyperbolic embedding preserves the intrinsic hierarchy of the EC system with theoretical guarantees in low-dimensional spaces. Extensive experiments on four datasets from the CARE benchmark demonstrate that PoinnCARE consistently and significantly outperforms state-of-the-art methods in EC number prediction.