ICLR 2026 论文列表

Accurate analysis of Medical time series (MedTS) data, such as Electroencephalography (EEG) and Electrocardiography (ECG), plays a pivotal role in healthcare applications, including the diagnosis of brain and heart diseases. MedTS data typically exhibits two critical patterns: **temporal dependencies** within individual channels and **channel dependencies** across multiple channels. While recent advances in deep learning have leveraged Transformer-based models to effectively capture temporal dependencies, they often struggle to model channel dependencies. This limitation stems from a structural mismatch: ***MedTS signals are inherently centralized, whereas the Transformer's attention is decentralized***, making it less effective at capturing global synchronization and unified waveform patterns. To bridge this gap, we propose **CoTAR** (Core Token Aggregation-Redistribution), a centralized MLP-based module tailored to replace the decentralized attention. Instead of allowing all tokens to interact directly, as in attention, CoTAR introduces a global core token that acts as a proxy to facilitate the inter-token interaction, thereby enforcing a centralized aggregation and redistribution strategy. This design not only better aligns with the centralized nature of MedTS signals but also reduces computational complexity from quadratic to linear. Experiments on five benchmarks validate the superiority of our method in both effectiveness and efficiency, achieving up to a **12.13%** improvement on the APAVA dataset, with merely 33% memory usage and 20% inference time compared to the previous state-of-the-art. Code and all training scripts are available in this [**Link**](https://github.com/Levi-Ackman/TeCh).

Post-training methods, especially Supervised Fine-Tuning (SFT) and Reinforcement Learning (RL), play an important role in improving large language models' (LLMs) complex reasoning abilities. However, the dominant two-stage pipeline (SFT then RL) suffers from a key inconsistency: SFT enforces rigid imitation that suppresses exploration and induces forgetting, limiting RL's potential for improvements. We address this inefficiency with TRAPO (**T**rust-**R**egion **A**daptive **P**olicy **O**ptimization), a hybrid framework that interleaves SFT and RL within each training instance by optimizing SFT loss on expert prefixes and RL loss on the model's own completions, unifying external supervision and self-exploration. To stabilize training, we introduce Trust-Region SFT (TrSFT), which minimizes forward KL divergence inside a trust region but attenuates optimization outside, effectively shifting toward reverse KL and yielding stable, mode-seeking updates favorable for RL. An adaptive prefix-selection mechanism further allocates expert guidance based on measured utility. Experiments on five mathematical reasoning benchmarks show that TRAPO consistently surpasses standard SFT, RL, and SFT-then-RL pipelines, as well as recent state-of-the-art approaches, establishing a strong new paradigm for reasoning-enhanced LLMs.

Recovering true signals from noisy measurements is a central challenge in inverse problems spanning medical imaging, geophysics, and signal processing. Current solutions nearly always balance prior assumptions regarding the true signal (regularization) with agreement to noisy measured data (data-fidelity). Conventional data-fidelity loss functions, such as mean-squared error (MSE) or negative log-likelihood, seek pointwise agreement with noisy measurements, often leading to overfitting to noise. In this work, we instead evaluate data-fidelity collectively by testing whether the observed measurements are statistically consistent with the noise distributions implied by the current estimate. We adopt this aggregated perspective and introduce $\textit{distributional consistency (DC) loss}$, a data-fidelity objective that replaces pointwise matching with distribution-level calibration. DC loss acts as a direct and practical plug-in replacement for standard data consistency terms: i) it is compatible with modern unsupervised regularizers that operate without paired measurement–ground-truth data, ii) it is optimized in the same way as traditional losses, and iii) it avoids overfitting to measurement noise without early stopping or the use of priors. Its scope naturally fits many practical inverse problems where the measurement-noise distribution is known and where the measured dataset consists of many independent noisy values. We demonstrate efficacy in two key example application areas: i) in image denoising with deep image prior, using DC instead of MSE loss removes the need for early stopping and achieves higher PSNR; ii) in medical image reconstruction from Poisson-noisy data, DC loss reduces artifacts in highly-iterated reconstructions and enhances the efficacy of hand-crafted regularization. These results position DC loss as a statistically grounded, performance-enhancing alternative to conventional fidelity losses for an important class of unsupervised noise-dominated inverse problems.

Offline Multi-Agent Reinforcement Learning (MARL) enables policy learning from static datasets in multi-agent systems, eliminating the need for risky or costly environment interactions during training. A central challenge in offline MARL lies in achieving effective collaboration among heterogeneous agents under the constraints of fixed datasets, where \textbf{conservatism} is introduced to restrict behaviors to data-supported distributions. Agents with distinct roles and capabilities require individualized conservatism - yet must maintain cohesive team performance. However, existing approaches often apply uniform conservatism across all agents, leading to over-constraining critical agents and under-constraining others, which hampers effective collaboration. To address this issue, a novel framework, \textbf{OMCDA}, is proposed, where the degree of conservatism is dynamically adjusted for individual agents based on their impact on overall system performance. The framework is characterized by two key innovations: (1) A decomposed Q-function architecture is introduced to disentangle return computation from policy deviation assessment, allowing precise evaluations of each agent's contribution; and (2) An adaptive conservatism mechanism is developed to scale constraint strength according to both behavior policy divergence and the estimated importance of agents to the system. Experiments on MuJoCo and SMAC show OMCDA outperforms existing offline MARL methods, effectively balancing the flexibility and conservatism across agents while ensuring fair credit assignment and better collaboration.

Reinforcement learning (RL) agents often face high sample complexity in sparse or delayed reward settings, due to limited prior knowledge. Conversely, large language models (LLMs) can provide subgoal structures, plausible trajectories, and abstract priors that support early learning. Yet heavy reliance on LLMs introduces scalability issues and risks dependence on unreliable signals, motivating ongoing efforts to integrate LLM guidance without compromising RL’s autonomy. We propose MIRA (Memory-Integrated Reinforcement Learning Agent), which incorporates a structured, evolving memory graph to guide early learning. This graph stores decision-relevant information, such as trajectory segments and subgoal decompositions, and is co-constructed from the agent’s high-return experiences and LLM outputs, amortizing LLM queries into a persistent memory instead of relying on continuous real-time supervision. From this structure, we derive a utility signal that softly adjusts advantage estimation to refine policy updates without altering the underlying reward function. As training progresses, the agent’s policy surpasses the initial LLM-derived priors, and the utility term decays, leaving long-term convergence guarantees intact. We show theoretically that this utility-based shaping improves early-stage learning in sparse-reward settings. Empirically, MIRA outperforms RL baselines and reaches returns comparable to methods that rely on frequent LLM supervision, while requiring substantially fewer online LLM queries.

We present Mixture-of-Length (MoL), an approach for Question Answering (QA) with context that aims to improve the balance between reasoning quality and response efficiency. Our method introduces a principled difficulty assessment based on information-theoretic principles and a dual-objective reward mechanism that adaptively modulates response length. In our experiments, MoL exhibits an emergent behavior termed "intelligent brevity": the model tends to produce shorter responses for simpler queries and longer ones for more complex inputs. This property is desirable for human-computer interaction and can reduce inference costs. A post-hoc analysis of internal activations suggests a correlation between this output adaptivity and the effective number of layers that contribute during inference. On multiple QA benchmarks, MoL demonstrates competitive accuracy while substantially reducing tokens compared to baselines, indicating that difficulty-aware length modulation is a promising direction for efficient QA with context.

Diffusion Language Models (DLMs) represent a promising alternative to autoregressive language models, using bidirectional masked token prediction. Yet their susceptibility to privacy leakage via Membership Inference Attacks (MIA) remains critically underexplored. This paper presents the first systematic investigation of MIA vulnerabilities in DLMs. Unlike the autoregressive models' single fixed prediction pattern, DLMs' multiple maskable configurations exponentially increase attack opportunities. This ability to probe many independent masks dramatically improves detection chances. To exploit this, we introduce SAMA (Subset-Aggregated Membership Attack), which addresses the sparse signal challenge through robust aggregation. SAMA samples masked subsets across progressive densities and applies sign-based statistics that remain effective despite heavy-tailed noise. Through inverse-weighted aggregation prioritizing sparse masks' cleaner signals, SAMA transforms sparse memorization detection into a robust voting mechanism. Experiments on nine datasets show SAMA achieves 30\% relative AUC improvement over the best baseline, with up to 8$\times$ improvement at low false positive rates. These findings reveal significant, previously unknown vulnerabilities in DLMs, necessitating the development of tailored privacy defenses.

Obtaining high-quality generations in modern LLMs has largely been framed as a selection problem: identifying a single winning generation from a diverse pool of $N$ samples, the Best-of-$N$ (BoN). Yet, this approach is inherently zero-sum, discarding diverse and potentially useful information from the pool. Instead, we explore a collaborative setup, where all candidates can potentially contribute to the final winning generation. To this end, we propose Fusion-of-$N$ (FusioN): a method that uses a general LLM judge to synthesize the most informative elements of each sample into a single final answer. We compare FusioN to BoN in two settings, (i) test-time scaling, where we sample and aggregate from a single model at test-time (ii) synthetic data generation, where we fuse samples from a pool of diverse teachers to improve a student model. We extensively benchmark both setups across 11 languages, 3 diverse benchmarks and varying model scales. Across the bench, FusionN consistently outperforms BoN showing versatility and robustness both in test-time scaling and in downstream gains from synthetic data generation. We also perform extensive analysis on FusioN, where it shows surprising strengths and robustness under challenging settings. These results show that we should shift how we think about evaluating and utilizing LLM generations from a monolithic measure of quality, to embracing their polylithic nature. This shift allows us to integrate diverse strengths, unlock latent potential, and achieve improvements that were previously inaccessible through selection alone.

Current foundation models for electroencephalography (EEG) rely on architectures adapted from computer vision or natural language processing, typically treating neural signals as pixel grids or token sequences. This approach overlooks that the neural activity is activated by diverse sparse coding across a complex geometric topological cortex. Inspired by biological neural mechanisms, we propose the Unified Neural Topological Foundation Model (Uni-NTFM), an architecture rooted in three core neuroscience principles. In detail, to align with the brain's decoupled coding mechanism, we design the Heterogeneous Feature Projection Module. This module simultaneously encodes both time-domain non-stationary transients and frequency-domain steady-state rhythms, ensuring high quality in both waveform morphology and spectral rhythms. Moreover, we introduce a Topological Embedding mechanism to inject structured spatial priors and align different sensor configurations onto a unified latent functional topography, effectively reconstructing the geometry of brain regions. Furthermore, we achieve functional modularization and sparse coding efficiency of biological networks by constructing the Mixture-of-Experts Transformer network. This dynamic routing mechanism assigns different signal patterns and tasks to specialized neural subnetworks, and effectively preventing task interference while increasing the model capacity to record-breaking 1.9 billion parameters. Uni-NTFM is pre-trained on a diverse corpus comprising 28,000 hours of EEG data, and outperforms existing models across nine distinct downstream tasks under both linear probing and fine-tuning settings, demonstrating that aligning model architecture with neural mechanisms is significant to learn universal representations and achieve generalizable brain decoding.} Our code is available at \url{https://anonymous.4open.science/r/Uni-NTFM-0924}

Abductive reasoning in knowledge graphs aims to generate plausible logical hypotheses from observed entities, with broad applications in areas such as clinical diagnosis and scientific discovery. However, due to a lack of controllability, a single observation may yield numerous plausible but redundant or irrelevant hypotheses on large-scale knowledge graphs. To address this limitation, we introduce the task of controllable hypothesis generation to improve the practical utility of abductive reasoning. This task faces two key challenges when controlling for generating long and complex logical hypotheses: hypothesis space collapse and hypothesis reward oversensitivity. To address these challenges, we propose **CtrlHGen**, a **C**on**tr**ollable **l**ogcial **H**ypothesis **Gen**eration framework for abductive reasoning over knowledge graphs, trained in a two-stage paradigm including supervised learning and subsequent reinforcement learning. To mitigate hypothesis space collapse, we design a dataset augmentation strategy based on sub-logical decomposition, enabling the model to learn complex logical structures by leveraging semantic patterns in simpler components. To address hypothesis reward oversensitivity, we incorporate smoothed semantic rewards including Dice and Overlap scores, and introduce a condition-adherence reward to guide the generation toward user-specified control constraints. Extensive experiments on three benchmark datasets demonstrate that our model not only better adheres to control conditions but also achieves superior semantic similarity performance compared to baselines. Our code is available at https://github.com/HKUST-KnowComp/CtrlHGen.

Recent advances in multimodal large language models (MLLMs) have catalyzed transformative progress in affective computing, enabling models to exhibit emergent emotional intelligence. Despite substantial methodological progress, current emotional benchmarks remain limited, as it is still unknown: (a) the generalization abilities of MLLMs across distinct scenarios, and (b) their reasoning capabilities to identify the triggering factors behind emotional states. To bridge these gaps, we present MME-Emotion, a systematic benchmark that assesses both emotional understanding and reasoning capabilities of MLLMs, enjoying scalable capacity, diverse settings, and unified protocols. As the largest emotional intelligence benchmark for MLLMs, MME-Emotion contains over 6,000 curated video clips with task-specific questioning-answering (QA) pairs, spanning broad scenarios to formulate eight emotional tasks. It further incorporates a holistic evaluation suite with hybrid metrics for emotion recognition and reasoning, analyzed through a multi-agent system framework. Through a rigorous evaluation of 20 advanced MLLMs, we uncover both their strengths and limitations, yielding several key insights: (1) Current MLLMs exhibit unsatisfactory emotional intelligence, with the best-performing model achieving only $39.3\%$ recognition score and $56.0\%$ Chain-of-Thought (CoT) score on our benchmark. (2) Generalist models (\emph{e.g.}, Gemini-2.5-Pro) derive emotional intelligence from generalized multimodal understanding capabilities, while specialist models (\emph{e.g.}, R1-Omni) can achieve comparable performance through domain-specific post-training adaptation. By introducing MME-Emotion, we hope that it can serve as a foundation for advancing MLLMs' emotional intelligence in the future.

The generation of quadrilateral-dominant meshes is a cornerstone of professional 3D content creation. However, existing generative models generate quad meshes by first generating triangle meshes and then merging triangles into quadrilaterals with some specific rules, which typically produces quad meshes with poor topology. In this paper, we introduce QuadGPT, the first autoregressive framework for generating quadrilateral meshes in an end-to-end manner. QuadGPT formulates this as a sequence prediction paradigm, distinguished by two key innovations: a unified tokenization method to handle mixed topologies of triangles and quadrilaterals, and a specialized Reinforcement Learning fine-tuning method tDPO for better generation quality. Extensive experiments demonstrate that QuadGPT significantly surpasses previous triangle-to-quad conversion pipelines in both geometric accuracy and topological quality. Our work establishes a new benchmark for native quad-mesh generation and showcases the power of combining large-scale autoregressive models with topology-aware RL refinement for creating structured 3D assets.

Retrofitting large language models (LLMs) with new behaviors typically requires full finetuning or distillation—costly steps that must be repeated for every architecture. In this work, we introduce ⌘V (Command-V), a backpropagation-free behavior transfer method that copies an existing residual representation adapter from a donor model and pastes its effect into an architecturally different recipient model. ⌘V profiles layer activations on a small prompt set, derives linear converters between corresponding layers, and applies the donor intervention in the recipient’s activation space. This process does not require access to the original training data and needs minimal compute. In three case studies—safety-refusal enhancement, jailbreak facilitation, and automatic chain-of-thought reasoning—⌘V matches the performance of direct finetuning while using orders of magnitude less resources.

The quadratic complexity of full attention mechanisms poses a significant bottleneck for Video Diffusion Models (VDMs) aiming to generate long-duration, high-resolution videos. While various sparse attention methods have been proposed, many are designed as training-free inference accelerators or do not optimally capture the unique spatio-temporal characteristics inherent in video data when trained natively. This paper introduces Video Mixture of Block Attention (VMoBA), a novel sparse attention mechanism specifically adapted for VDMs. Motivated by an in-depth analysis of attention patterns within pre-trained video transformers, which revealed strong spatio-temporal locality, varying query importance, and head-specific concentration levels, VMoBA enhances the original MoBA framework with three key modifications: (1) a layer-wise recurrent block partition scheme (1D-2D-3D) to dynamically adapt to diverse spatio-temporal attention patterns and improve efficiency; (2) global block selection to prioritize the most salient query-key block interactions across an entire attention head; and (3) threshold-based block selection to dynamically determine the number of attended blocks based on their cumulative similarity. Extensive experiments demonstrate that VMoBA significantly accelerates the training of VDMs on longer sequences, achieving 2.92$\times$ FLOPs and 1.48$\times$ latency speedup, while attaining comparable or even superior generation quality to full attention. Furthermore, VMoBA exhibits competitive performance in training-free inference, offering 2.40$\times$ FLOPs and 1.35$\times$ latency speedup for high-res video generation.

While humans naturally learn and adapt from past experiences, large language models (LLMs) and their agentic counterparts often fail to retain reasoning from previous tasks and apply it in future contexts. We introduce **L**og-**A**ugmented **G**eneration (LAG), a novel framework that *directly reuses* prior computation and reasoning from past logs at test time, enabling models to learn from previous tasks and perform better on new, unseen challenges, without sacrificing the system's efficiency or scalability. Our approach represents task logs as key-value (KV) caches that encode the full reasoning context of prior tasks, while storing KV values for only a selected subset of tokens. When a new task arises, LAG retrieves KV values from relevant logs to augment generation. Unlike reflection-based memory mechanisms, which require additional extraction or distillation steps, LAG reuses prior reasoning verbatim. Moreover, it extends beyond existing KV caching techniques, which have primarily targeted efficiency, by explicitly improving accuracy through log reuse. Experiments on knowledge- and reasoning-intensive datasets demonstrate that our method significantly outperforms standard agentic systems that do not utilize logs, as well as existing solutions based on reflection and KV cache techniques.

Reward models (RMs) are fundamental to aligning Large Language Models (LLMs) via human feedback, yet they often suffer from reward hacking. They tend to latch on to superficial or spurious attributes, such as response length or formatting, mistaking these cues learned from correlations in training data for the true causal drivers of quality (e.g., factuality, relevance). This occurs because standard training objectives struggle to disentangle these factors, leading to brittle RMs and misaligned policies. We introduce CROME (Causally Robust Reward Modeling), a novel framework inspired by an explicit causal model designed to mitigate reward hacking. CROME queries an oracle LLM for rubrics that are (or the oracle deems to be) causally relevant to answering a specific prompt. Then, it employs the following synthetic targeted augmentations during training: (1) Causal Augmentations, which are pairs that differ along specific causal attributes (subset of the Oracle identified rubrics), to enforce sensitivity along each causal attribute individually, and (2) Neutral Augmentations, which are tie-label pairs varying primarily in spurious attributes, to enforce invariance along spurious attributes. Notably, our neutral augmentations are produced without any knowledge of unknown spurious factors, via question swapping and response interventions only along causal rubrics. We show that the CROME augmentation strategy using rubrics from popular LLM APIs significantly outperforms standard baselines on RewardBench, improving average accuracy by up to 5.3% and achieving gains of up to 7.1% and 12.4% in reasoning and safety. The robustness of CROME is further testified by significant gains in DPO-aligned policies and Best-of-N alignment across various benchmarks, including AlpacaEval 2.0, RewardBench, safety-focused WildGuardTest, and the reasoning-specific GSM8k.

Learning on text-attributed graphs has motivated the use of Large Language Models (LLMs) for graph learning. However, most fusion strategies are applied uniformly across all nodes and attain only small overall performance gains. We argue this result stems from aggregate metrics that obscure when LLMs provide benefit, inhibiting actionable signals for new strategies. In this work, we reframe LLM–GNN fusion around nodes where GNNs typically falter. We first show that performance can significantly differ between GNNs and LLMs, with each excelling on distinct structural patterns, such as local homophily. To leverage this finding, we propose **GLANCE** (**G**NN with **L**LM **A**ssistance for **N**eighbor- and **C**ontext-aware **E**mbeddings), a framework that invokes an LLM to refine a GNN's prediction. GLANCE employs a lightweight router that, given inexpensive per-node signals, decides whether to query the LLM. Since the LLM calls are non-differentiable, the router is trained with an advantage-based objective that compares the utility of querying the LLM against relying solely on the GNN. Across multiple benchmarks, GLANCE achieves the best performance balance across node subgroups, achieving significant gains on heterophilous nodes (up to +5.8%) while simultaneously achieving top overall performance (up to +1.1%). Our findings advocate for adaptive, node-aware, GNN-LLM architectures, showing that selectively invoking the LLM where it adds value enables scalable application of LLMs to large graphs.

Efficient code retrieval is critical for developer productivity, yet existing benchmarks largely focus on Python and rarely stress-test robustness beyond superficial lexical cues. To address the gap, we introduce an automated pipeline for code search datasets and present CLARC, a C/C++ benchmark built from real-world GitHub repositories. CLARC contains 1,245 query-code pairs for evaluation and 5,472 pairs for training. The benchmark incorporates LLM-generated natural language queries validated through rigorous human scoring and hypothesis testing. To analyze contextual requirements effectively, our pipeline starts by ensuring code compilability. It then categorizes code snippets by dependency complexity, distinguishing whether the code relies on custom-defined types or helper functions. The pipeline also enables CLARC to stress-test retrieval robustness by introducing challenging settings, including identifier anonymization and compilation to low-level languages like Assembly and WebAssembly. Under these conditions, our evaluation of six state-of-the-art models reveals sharp drops in retrieval effectiveness. The experimental results highlight the models' persistent reliance on lexical features rather than code semantic understanding. Our dataset is publicly available at https://huggingface.co/datasets/ClarcTeam/CLARC.

Reinforcement learning (RL) with verifiable rewards has recently catalyzed a wave of “MLLM-r1” approaches that bring RL to vision language models. Most representative paradigms begin with a cold start, typically employing supervised fine-tuning (SFT), to initialize the policy before RL. However, SFT-based cold start adopts the reasoning paradigm intertwined with task solution and output format, which may induce instruction-style overfitting, weakens out-of-distribution generalization, and ultimately affects downstream RL. We revisit the cold start along two views, its training method and data construction, and introduce the Generalization Factor (GF) coefficient to quantify the generalization capability under different methods. Our empirical study finds that preference–based training methods (e.g. DPO) generalizes better than SFT-based methods in cold start. Motivated by this, we propose $\textbf{SPECS}$—a $\textbf{S}$elf-distilled, $\textbf{P}$r$\textbf{e}$ference-based $\textbf{C}$old $\textbf{S}$tart framework that decouples multimodal learning: (1) generates introspective preference data pairs via self-distillation, avoiding reliance on larger teachers or manual annotation; (2) performs preference–based training to learn, focusing on shallow, transferable surface-form criteria (format, structure, style) rather than memorizing content; and (3) hands off to RL with verifiable rewards for deep reasoning results. Experimental results across multiple multimodal benchmarks show that our decoupling learning framework yields consistent performance gains over strong baselines, improving MEGA-Bench by 4.1\% and MathVista by 12.2\%. Additional experiments indicate that SPECS contributes to reducing in-distribution “stuckness,” improving exploration, stabilizing training, and raising the performance ceiling.