ICML 2026 论文列表

Conformal prediction converts any black-box predictor into one with finite-sample, distribution-free coverage guarantees, outputting prediction sets $T(x)$ that contain the true label with probability at least $1-\alpha$. To construct these prediction sets, conformal prediction relies on a randomly sampled ``calibration set'' of labeled examples. In many applications, however, this labeled calibration set is costly to collect, creating a tradeoff between upfront labeling cost and downstream utility of the conformal predictor. In this work, we study *conformal prediction with costly label queries*, where unlabeled examples arrive i.i.d. and labels can be queried one at a time. After $m$ queries, we form a conformal predictor; the upfront cost of this predictor is the calibration set size $m$, and its efficiency is the expected prediction set size $\mathbb{E}|T_m(X)|$. We design an online stopping rule $\hat{m}$ that automatically balances the upfront cost against conformal efficiency *while preserving the original conformal guarantee*. Theoretically, we show that under mild regularity assumptions, the expected total cost of our stopping rule matches the best fixed calibration size in hindsight. Experimentally, we find that our stopping rule reduces cost compared to standard choices of $m$ from the literature by 40.6% $\pm$ 2.3%. Finally, we demonstrate a reduction from the probably approximately correct labeling problem of Candès et al. (2025) to CP, under which our stopping rule minimizes the total labeling cost.

Safety alignment in large language models is typically evaluated under isolated queries, yet real-world use is inherently multi-turn. Although multi-turn jailbreaks are empirically effective, the structure of conversational safety failure remains insufficiently understood. In this work, we study safety failures from a state-space perspective and show that many multi-turn safety failures in current safety-aligned language models arise from contextual state evolution, a regime that is not fully captured by isolated prompt-level analyses alone. We introduce STAR, a state-oriented diagnostic framework that treats dialogue history as a state transition operator and enables controlled analysis of safety behavior along interaction trajectories. Rather than optimizing attack strength, STAR provides a principled probe of how aligned models traverse the safety boundary under autoregressive conditioning. Across multiple frontier language models, we find that systems which appear robust under static evaluation can undergo rapid and reproducible safety collapse under structured multi-turn interaction. Mechanistic analysis reveals monotonic drift away from refusal-related representations and abrupt phase transitions induced by role-conditioned context. Together, these findings motivate viewing language model safety as a dynamic, state-dependent process defined over conversational trajectories.

Monocular 3D human pose estimation is fundamentally challenged by severe occlusion and inherent depth ambiguity. To address this, we propose Glimpse, a framework that learns robust 3D poses by explicitly modeling anatomical geometry from a single image. We recast the problem as geometry learning of multi-scale structural priors, realized through two synergistic components. First, structured sampling captures the body's geometric continuity through dual-level feature extraction, acquiring both local joint appearance and continuous features along skeletal limbs via deformable sampling. By propagating limb-level geometric cues to their connected joints, this design bridges information gaps caused by occlusion. Second, geometric correction ensures global 3D consistency by lifting coherent 2D features into a canonical 3D reference space, where a shared 3D anchor guides a distance-aware fusion mechanism. Extensive experiments conducted on Human3.6M and MPI-INF-3DHP demonstrate that Glimpse achieves state-of-the-art performance, with superior robustness under severe occlusion and complex articulation.

Personalization techniques for image generation models have increasingly been misused for malicious purposes, including unauthorized style imitation and copyrighted content replication. In response, recent mimicry protection methods embed carefully designed perturbations into images to disrupt a model’s ability to learn genuine semantic representations. Despite their growing adoption, the robustness of these protection mechanisms remains poorly understood, raising concerns about their reliability in real-world deployment. In this work, we present the first systematic analysis showing that DDIM inversion inherently acts as a perturbation amplifier, causing protected images to suffer severe structural distortions during reconstruction. Building on this observation, we propose DDIM Inversion-based Reconstruction Purification (DIRP), a novel purification approach that removes protective perturbations by explicitly minimizing DDIM inversion reconstruction error under perceptual constraints. Extensive experiments on six existing mimicry protection methods demonstrate that DIRP consistently outperforms five state-of-the-art attack baselines, achieving superior perturbation removal while better preserving image quality. Our results expose fundamental vulnerabilities in current mimicry protection strategies and highlight the urgent need for more robust and principled defenses.

Clustering is commonly compared through leaderboards that collapse performance into a single aggregate ranking. Such summaries obscure why methods succeed, which data properties align with failure, and how conclusions shift under representation changes and realistic tuning constraints. We present CHB, a diagnostic toolkit for hardness-aware clustering evaluation. CHB maps each dataset--representation pair to an interpretable hardness fingerprint capturing (i) separation, (ii) cohesion and scale heterogeneity, and (iii) topology through scalable persistent-homology summaries. Using this diagnostic space, CHB evaluates clustering algorithms under standardized, compute-aware tracks. Conditioning results on hardness coordinates turns comparison into diagnosis: across a broad range of datasets and their representations, CHB reveals reproducible structural regimes, uncovers regime-dependent ranking reversals across method families, and surfaces robustness signatures, including topology-linked breakdowns. CHB further enables representation auditing by attributing gains to measurable shifts in the hardness fingerprint rather than just external performance changes. We release CHB as an open, extensible artifact for evaluating new clustering methods and embeddings within a shared diagnostic framework.

Foundation models have revolutionized AI, yet biological applications often repurpose general architectures without accounting for the intrinsic structural and functional properties of distinct modalities, such as genomic and proteomic sequences. Consequently, these architectures lack the inductive biases required to capture the complex ``grammars" inherent to biological data, resulting in suboptimal performance. To address this, we introduce BioArc, a framework utilizing Neural Architecture Search (NAS) to shift from intuition-driven design to automated data-driven discovery. Unlike standard NAS restricted to homogeneous spaces, BioArc navigates a heterogeneous space for open-ended composition of architectural blocks. By systematically analyzing the interplay between architecture, tokenization, and training across modalities, BioArc identifies novel hybrid architectures that surpass state-of-the-art models while being up to 25x smaller. We distill these findings into empirical design principles and validate their biological relevance, demonstrating how our designs hierarchically capture the underlying biological grammar. Additionally, we introduce an agentic framework to predict optimal architectures for new tasks. Overall, BioArc provides a data-driven methodology for developing the next generation of efficient biological foundation models and task-specific networks.

Quantifying the distance between datasets is a fundamental question in mathematics and machine learning. We propose *magnitude distance*, a novel distance metric defined on finite datasets using the notion of the *magnitude* of a metric space. The proposed distance incorporates a tunable scaling parameter, $t$, that controls the sensitivity to global structure (small $t$) and finer details (large $t$). We prove several theoretical properties of magnitude distance, including its limiting behavior across scales and conditions under which it satisfies key metric properties. In contrast to classical distances, we show that magnitude distance remains discriminative in high-dimensional settings when the scale is appropriately tuned. We further demonstrate how magnitude distance can be used as a training objective for push-forward generative models. Our experimental results support our theoretical analysis and demonstrate that magnitude distance provides meaningful signals, comparable to established distance-based generative approaches.

Reinforcement learning from human feedback is the leading approach to aligning powerful AI systems so that they can be safe and helpful for humanity. While RLHF is typically modelled as a problem of learning a single preference ranking from noisy feedback, true human preferences are complex and often conflicting, representing substantive disagreements stemming from the diversity of individual human values. With this motivation, a recent line of research has studied RLHF from the perspective of social choice theory, which provides a set of well-established desirable properties for aggregating diverse preferences. Seen through this lens, the standard learning objective in RLHF is equivalent to aggregating diverse human preferences via the Borda count rule. At the same time, several new RLHF algorithms have been proposed, which turn out to be equivalent to the von Neumann winner social choice rule. However, the connection between social choice theory and RLHF has thus far ignored the critical role of regularization to prevent divergence from a reference policy, which is utilized in essentially all practical RLHF algorithms. In this paper, we study how regularization affects the social choice axioms satisfied by different RLHF algorithms, and prove that regularization improves the axiomatic properties of the von Neumann winner rule. In contrast, the Borda count rule still fails to satisfy key social choice axioms even when regularized. These results provide a principled argument grounded in social choice theory for utilizing practical RLHF algorithms that correspond to the von Neumann winner, rather than the standard RLHF objective.

By sharing intermediate features, collaborative perception extends each agent's sensing beyond standalone limits, but real-world feature modality heterogeneity remains a key barrier to effective fusion. Most existing methods, including direct adaption and protocol-based transforma-tion, typically rely on training adapters for newly emerging feature modalities and often require additional retraining or fine-tuning. Such repeated training is costly and is often infeasible across manufacturers due to model and data privacy con-straints, limiting real-world scalability. To address this issue, we propose UniTrans, a universal any-to-any feature modality translation model that instantiates translators on the fly for arbitrary modalities. UniTrans pretrains a bank of translator expert parameters and learns their combination coefficients as a function of source-to-target modality mapping. The mapping is measured in a modality-intrinsic latent space, where an intrinsic encoder extracts modality-specific yet scene-invariant codes from single-frame intermediate features, enabling UniTrans to instantiate translators in a zero-shot manner. Experiments on OPV2V-H and DAIR-V2X demonstrate that UniTrans consistently outperforms state-of-the-art methods in both simulated and real-world set-tings, enabling efficient any-to-any translation through a universal model. Code will be made available.

Existing time series foundation models (TSFMs), often based on transformer variants, lack adaptability to different sampling rates, struggle with generalization across varying context and target lengths and are computationally inefficient. We introduce FlowState, a novel TSFM architecture that achieves sampling-rate–equivariant forecasting through a unified design that pairs a state space model (SSM) encoder with a functional basis decoder (FBD). This design enables continuous-time modeling and dynamic time-scale adjustment, allowing FlowState to inherently generalize across all possible temporal resolutions, and dynamically adjust the forecasting horizons without retraining. We further propose an efficient pretraining strategy that improves robustness and accelerates training. Despite being one of the smallest TSFMs, FlowState achieves state-of-the-art results on the widely used GIFT-Eval benchmark, while demonstrating superior adaptability to unseen sampling rates. Our detailed analyses confirm the effectiveness of its components, and we demonstrate its unique ability to adapt to varying input sampling rates.

Diffusion Language Models (DLMs) generate text by iteratively denoising a masked sequence, repeatedly deciding which positions to commit at each step. Standard decoding follows a greedy rule, unmasking the most confident positions, yet this local choice can lock the model into a suboptimal unmasking order, especially on reasoning-heavy prompts. We present Search Or AcceleRate (SOAR), a training-free decoding algorithm that adapts its behavior to the model’s uncertainty. When confidence is low, SOAR briefly widens the search over alternative unmasking decisions to avoid premature commitments; when confidence is high, it collapses the search and decodes many positions in parallel to reduce the number of denoising iterations. Across mathematical reasoning and code generation benchmarks (GSM8K, MBPP, HumanEval) on Dream-7B and LLaDA-8B, SOAR improves generation quality while maintaining competitive inference speed, offering a practical way to balance quality and efficiency in DLM decoding.

We present SleepLM, a family of sleep-language foundation models that enable human sleep alignment, interpretation, and interaction with natural language. Despite the critical role of sleep, learning-based sleep analysis systems operate in closed label spaces (e.g., predefined stages or events) and fail to describe, query, or generalize to novel sleep phenomena. SleepLM bridges natural language and multimodal polysomnography, enabling language-grounded representations of sleep physiology. To support this alignment, we introduce a multilevel sleep caption generation pipeline that enables the curation of the first large-scale sleep-text dataset, comprising over 100K hours of data from more than 10,000 individuals. Furthermore, we present a unified pretraining objective that combines contrastive alignment, caption generation, and signal reconstruction to better capture physiological fidelity and cross-modal interactions. Extensive experiments on real-world sleep understanding tasks verify that SleepLM outperforms state-of-the-art in zero-shot and few-shot learning, cross-modal retrieval, and sleep captioning. Importantly, SleepLM also exhibits intriguing capabilities including language-guided event localization, targeted insight generation, and zero-shot generalization to unseen tasks. To support reproducibility and future work, we open-source the captioning pipeline, pretrained checkpoints, and the model architectures at https://github.com/yang-ai-lab/SleepLM.

Linear attention mitigates the quadratic complexity of softmax attention but suffers from a critical loss of expressiveness. We identify two primary causes: (1) The normalization operation cancels the query norm, which breaks the correlation between a query's norm and the spikiness (entropy) of the attention distribution as in softmax attention. (2) Standard techniques for enforcing non-negativity cause destructive information loss by nullifying valid inner-product interactions. To address these challenges, we introduce **NaLaFormer**, a novel linear attention mechanism built upon a norm$\times$direction (ND) decomposition of the query and key vectors. We leverage each component to solve a distinct problem: The *query norm* is injected into our kernel to create a query-norm-aware map that restores the attention distribution's spikiness. The *direction vectors* are processed by a geometric, cosine-based similarity metric that guarantees non-negativity while preserving the rich, fine-grained information of the inner product. We validate NaLaFormer through a comprehensive multi-modal evaluation, where it sets new state-of-the-art benchmarks for linear attention. Our model achieves up to a 7.5\% accuracy gain on ImageNet-1K and a 4.7\% mIoU improvement on ADE20K over comparable baselines. It demonstrates profound efficiency, reducing peak memory by a transformative 92.3\% in token-intensive super-resolution tasks (70K+ tokens). NaLaFormer's versatility is further confirmed as it surpasses strong baselines like Mamba on common-sense reasoning and sets a new state-of-the-art on the Long Range Arena (LRA) benchmark. Source code can be found in the supplementary materials.

Robots in dynamic, human-centric environments must follow language instructions while maintaining real-time reactive control. Vision-language-action (VLA) models offer a promising framework, but they assume temporally aligned reasoning and control, despite semantic inference being inherently delayed relative to real-time action. We introduce Think-in-Control (TIC)-VLA, a latency-aware framework that explicitly models delayed semantic reasoning during action generation. TIC-VLA defines a delayed semantic-control interface that conditions action generation on delayed vision-language semantic states and explicit latency metadata, in addition to current observations. We further propose a latency-consistent training pipeline that injects reasoning inference delays during imitation learning and online reinforcement learning, aligning training with asynchronous deployment. To support realistic evaluation, we present DynaNav, a physics-accurate, photo-realistic simulation suite for language-guided navigation in dynamic environments. Extensive experiments in simulation and on a real robot show that TIC-VLA consistently outperforms prior VLA models while maintaining robust real-time control under multi-second reasoning latency. Project website: https://ucla-mobility.github.io/TIC-VLA/

Many machine learning problems are geometric at their core, relying on metric representations of data for tasks such as clustering, prototype selection, nearest-neighbor search, and graph-based learning. Furthermore, data is constantly evolving and it is routinely transformed through dimensionality reduction, random projections, feature embeddings, compression, or privacy-preserving mechanisms. These transformations are designed to preserve geometry approximately. As a result, they preserve objective values for many geometric optimization problems, but they fail to guarantee that algorithmic outcomes remain consistent. In this work, we study \emph{resilient data summaries} for geometric optimization. Building on the notion of \emph{$\gamma$-resilient algorithms} from Ahmadian, we introduce $\gamma$-resilient coresets. A $\gamma$-resilient $(k,\varepsilon)$-coreset is a compact, weighted summary that guarantees a $(1+\varepsilon)$ approximation to the objective and enforces stability at the level of assignments. We complement our positive result with a lower bound showing that to obtain a tight approximation for resilient clustering it is necessary to use a bi-criteria solution.

Supervised Fine-Tuning with Negative Log-Likelihood (NLL) remains the standard post-training paradigm for Large Language Models, yet it imposes a disproportionately aggressive update force on low-probability target tokens. This focus forces the model to prioritize minimizing the loss of difficult samples over optimizing the overall quality of the generation, often leading to unwarranted overconfidence. On the other hand, alternatives like Dynamic Fine-Tuning suffer from vanishing gradients on these tokens, which severely hinders the acquisition of new concepts. To bridge this gap, we propose **S**pectrum-**A**daptive **F**ine-**T**uning (**SAFT**), a unified framework that interpolates between the aggressive learning signal of NLL and the robust nature of probability-weighted optimization. By adaptively balancing these objectives, SAFT effectively mitigates outlier sensitivity without sacrificing learning efficiency. Empirically, our method achieves state-of-the-art performance on mathematical reasoning benchmarks, demonstrating superior generalization on out-of-distribution tasks. Furthermore, evaluations on general conversational alignment validate SAFT's broad adaptability across diverse data regimes. Our code is available at https://github.com/sjtu-scx/SAFT.

Multimodal systems encounter partial observability from sensor dropout and degradation, yet adverse tails can remain unreliable even when average performance is preserved. Under a locked same-support contract, matched-average runs can exhibit same-support tail failure, diverging on worst-case and shift-reweighted metrics over the same observable support. The failure arises because conditional interaction separates environment coverage from parameter exposure. Gated routing sends information through different groups, so high-leverage groups may receive few updates despite complete coverage, and tail aggregation amplifies their errors. We audit this mismatch with TailPressure, an exposure-normalized statistic of tail-leveraged interaction from gating logs. We propose Heterogeneity-aware Closed-loop Exposure Stabilizer (H-CES), a lightweight controller that stabilizes per-group pressure via deterministic increment-branch gating and group-wise decoupled weight decay, without changing loss or inference. Across multimodal settings and backbones, H-CES improves same-support tail reliability while preserving clean performance.

The abundance of pre-trained diffusion models provides an opportunity for composition. Combining several models, however, runs the risk of one model dominating or models disagreeing with each other. Here, we propose Divide-and-Denoise, a method for coordinating multiple pre-trained diffusion models during sampling. Much like managing a specialized workforce, our method creates a fair but efficient division of labor across models. Central to our method is the notion of an allocation which defines the responsibility of each model to every region of the noisy sample. At every timestep, we then denoise by (i) updating the allocation by solving a fair division game, where we divide the sample into regions that maximize total utility under fairness constraints, and (ii) aligning the models with this allocation, where we guide each model to denoise within its assigned region. This leads to a new composite denoising process that evolves in tandem with a division process. We evaluate Divide-and-Denoise on conditional image generation. Across several quality metrics, including the GenEval benchmark, our method outperforms baselines and resolves common failures including missing objects and mismatched attributes. Experiments show that Divide-and-Denoise utilizes each model's expertise without neglecting any other model.

Multimodal understanding of advertising videos is essential for interpreting the intricate relationship between visual storytelling and abstract persuasion strategies. However, despite excelling at general search, existing agents often struggle to bridge the cognitive gap between pixel-level perception and high-level marketing logic. To address this challenge, we introduce **AD-MIR**, a framework designed to decode advertising intent via a two-stage architecture. First, in the **Structure-Aware Memory Construction** phase, the system converts raw video into a structured database by integrating semantic retrieval with exact keyword matching. This approach prioritizes fine-grained brand details, such as logos and on-screen text, while dynamically filtering out irrelevant background noise to isolate key protagonists. Second, the **Structured Reasoning Agent** mimics a marketing expert through an iterative inquiry loop, decomposing the narrative to deduce implicit persuasion tactics. Crucially, it employs an evidence-based self-correction mechanism that rigorously validates these insights against specific video frames, automatically backtracking when visual support is lacking. Evaluation on the AdsQA benchmark demonstrates that AD-MIR achieves state-of-the-art performance, surpassing the strongest general-purpose agent, DVD, by 1.8 and 9.5 percentage points in strict and relaxed accuracy, respectively. These results underscore that effective advertising understanding demands explicitly grounding abstract marketing strategies in pixel-level evidence. The code is available at [https://github.com/Little-Fridge/AD-MIR](https://github.com/Little-Fridge/AD-MIR).