论文检索

Social Aspects · Alignment

Yangneng Chen, Jing Li

Large Vision-Language Models (LVLMs) extend large language models with visual understanding, but remain vulnerable to hallucination, where outputs are fluent yet inconsistent with images. Recent studies link this issue to language bias—the tendency of LVLMs to over-rely on text while neglecting visual inputs. Yet most analyses remain empirical without uncovering its underlying cause. In this paper, we provide a systematic study of language bias and identify its root in modality misalignment during training. Our analysis shows that both Visual Instruction Tuning (VIT) and Direct Preference Optimization (DPO) often prioritize textual improvements, which may cause LVLMs to overly lean toward language modeling rather than balanced multimodal understanding. To address this, we propose two simple yet effective methods: Language Bias Regularization (LBR), which mitigates language bias through regularization during instruction tuning, and Language Bias Penalty (LBP), which penalizes language bias in the DPO training process. Extensive experiments across diverse models and benchmarks demonstrate the effectiveness of our approach. LBR consistently improves performance on over ten general benchmarks, while LBP significantly reduces hallucination and improves trustworthiness. Together, these methods not only mitigate language bias but also advance the overall alignment of LVLMs, all without introducing any additional data or auxiliary models.

Theo X. Olausson, Joao Monteiro, Michal Klein, Marco Cuturi

Maximum inner product search (MIPS) is a crucial subroutine in machine learning, requiring identification of database vectors that align most strongly with a given query. We propose amortized MIPS: a learning-based approach that trains neural networks to directly predict MIPS solutions, amortizing the computational cost of search across queries drawn from a known distribution. Our key insight is that the MIPS value function - the maximum inner product as a function of the query - is convex (as the pointwise maximum of linear functions), and its gradient at each query equals the optimal database vector. We explore two complementary architectures: (1) Input Convex Neural Networks (ICNNs) that learn the convex value function and recover the optimal match via gradient computation, and (2) VectorICNNs that directly regress the argmax, bypassing gradient computation entirely at inference time. For ICNNs, we combine score regression with gradient matching losses; for VectorICNNs, we introduce a score consistency loss derived from Euler's theorem for homogeneous functions. We further propose homogenization wrappers that enforce positive 1-homogeneity, theoretically linking function values to gradients. Our experiments on retrieval benchmarks demonstrate that convexity provides an effective inductive bias, with learned potentials achieving high match rates while requiring only a single forward pass at inference.

Applications · Robotics

Yan Zhang, Zheng WANG, Pengpeng Zeng, Xing Xu, Jingkuan Song, Heng Tao Shen

Visuo-tactile sensors have been widely adopted in robotic manipulation. However, inherent heterogeneity in sensor designs hinders the learning of unified tactile representations in cross-sensor scenarios. Existing methods that focus on reconstruction or task-specific supervision often fail to capture the common information between different tactile sensors, particularly in the presence of substantial sensor variations, resulting in limited generalization to unseen sensors. To address this, we propose Cross-Tactile Sensor Representation Learning (CTSRL), a unified framework for sensor-agnostic tactile representation learning. CTSRL introduces a Cross-Sensor Modulator (CSM) to eliminate sensor-specific biases and adopts a two-stage learning paradigm: (1) leveraging aligned synthetic data for cross-sensor self-supervised learning to extract shared latent representations across sensor domains; and (2) integrating real-world multimodal tactile data to bridge the sim-to-real semantic gap through cross-modal alignment, thereby enriching representations with fine-grained semantic attributes. Experimental results show that our method demonstrates strong multi-sensor generalization, significantly improving sensor-agnostic representation learning.

Theory · Game Theory

Matteo Bollini, Matteo Castiglioni, Alberto Marchesi

*Hidden-action principal-agent problems* model scenarios in which a principal induces an agent to take a costly and *unobservable* action through the provision of outcome-dependent payments. These problems find application in a variety of real-world settings, such as crowdsourcing, online labor platforms, and machine learning task delegation. Recently, much of the literature has focused on how to handle the principal’s *uncertainty* about the agent and the surrounding environment, which is often the main challenge in practice. One prominent approach is to adopt an *online learning* framework, where the principal repeatedly interacts with the agent to learn optimal payments from experience. However, existing learning algorithms, while achieving regret that scales sublinearly in the number of interaction rounds $T$, typically suffer from an exponential dependence on the size of the problem instance. In this paper, we show that this problematic exponential growth can be avoided by assuming that the principal has knowledge of a set of possible actions of the agent, while remaining unaware of which actions are actually available---an assumption that is reasonable in many real-world settings.

General Machine Learning · Clustering

Zhihao Yao, Yuxuan Gu, Jixuan Yin, Bo Li

Pseudo-labeling based on Optimal Transport (OT) has become an effective mechanism for enhancing short text clustering. Existing OT methods are short in modeling semantic consistencies between samples, which may assign different pseudo-labels to semantically similar samples. These erroneous pseudo-labels can cause the model to produce inferior clusters. This paper proposes a novel short text clustering framework, which remedies the neglect of semantic consistency in existing OT methods, generating reliable pseudo-labels to facilitate clustering. Specifically, our method first proposes a novel instance-level attention mechanism to capture semantic relationships between samples, which are then integrated into the OT formulation to endow the transport process with neighborhood semantic awareness. By solving the proposed OT formulation, reliable pseudo-labels are obtained that simultaneously account for sample-to-sample semantic consistency and sample-to-cluster global structure information. These reliable pseudo-labels are then used as supervisory signals to guide the model to achieve accurate clustering. Extensive experiments demonstrate that the proposed method outperforms state-of-the-art approaches. The code is available at: https://anonymous.4open.science/r/RPDC-STC-8B53/README.md

Social Aspects · Alignment

Keshav Shenoy, Li Yang, Abhay Sheshadri, Jack Lindsey, Samuel Marks, Rowan Wang

Can we train LLMs to *introspect*, i.e. to faithfully describe their own behaviors in natural language? Prior work has shown some, limited, success. However, it is difficult to scale introspection training due to a lack of ground-truth labels. In this work, we study an approach to introspection training which side-steps this data bottleneck. Given a target model $M$, our method works by fine-tuning models $M_i$ from $M$ with implanted behaviors $b_i$ (such as downplaying medical problems); the $(M_i, b_i)$ pairs serve as labeled introspection training data. We then train an *introspection adapter* (IA): a LoRA adapter jointly optimized across the fine-tunes $M_i$ which causes them to verbalize their implanted behaviors. This IA induces faithful introspection in fine-tunes of $M$ that were trained in very different ways from the $M_i$, as well as in $M$ itself. This is surprising because the IA was never trained on $M$. To demonstrate the utility of IAs, we use them to successfully audit misaligned models introduced in prior work. IAs can also be used to detect fine-tuning API attacks which train models to comply with encrypted harmful requests. Notably, IAs are more effective when applied to larger models. Overall, our results suggest that IAs are a scalable, effective, and practically useful approach to LLM introspection training.

Deep Learning · Large Language Models

Geert Heyman, Frederik Vandeputte

Large language models can be steered at inference time through prompting or activation interventions, but activation steering methods often underperform compared to prompt-based approaches. We investigate whether activation steering can be improved by learning to mimic the interventions that prompt steering triggers within the model. To this end, we introduce *Prompt Steering Replacement (PSR)* models, a new family of activation steering methods that distill prompt steering behavior into interpretable interventions on model activations. A PSR is an activation steering method that estimates position-specific steering coefficients and is trained to imitate prompt-based interventions. Experiments on persona steering and instruction following across multiple language models demonstrate that PSR models consistently outperform constant-coefficient interventions that are frequently used in the literature and achieve performance close to or exceeding prompt steering while maintaining interpretability.

Reinforcement Learning · Deep RL

Guojian Zhan, Likun Wang, Pengcheng Wang, Feihong Zhang, Jingliang Duan, Kaicheng Yu, Masayoshi Tomizuka, Shengbo Li

Maximum entropy has become a mainstream off-policy reinforcement learning (RL) framework for balancing exploitation and exploration. However, two bottlenecks still limit further performance gains: (1) non-stationary Q-value estimation stemming from the joint injection of entropy and the concurrent updating of its temperature parameter; and (2) short-sighted local entropy tuning, which adjusts temperature solely based on current single-step entropy without accounting for cumulative entropy over time. In this paper, we broaden the maximum entropy framework by proposing a trajectory entropy-constrained reinforcement learning (TECRL) framework to address these limitations. We begin by introducing reward-entropy separation (RES) to decouple the value targets, ensuring they remain stable and unaffected by temperature fluctuations. Subsequently, the resulting entropy Q-function is leveraged to explicitly quantify expected cumulative entropy, allowing for the enforcement of a trajectory entropy constraint (TEC) to govern long-term stochasticity. We instantiate this framework as DSAC-E, a practical off-policy algorithm that builds upon the latest distributional soft actor-critic. Extensive evaluations across 10 challenging tasks in locomotion, robotic manipulation, and vision-based driving domains demonstrate that DSAC-E consistently outperforms baselines in both cumulative returns and training stability.

Social Aspects · Security

Pragati Meshram, Varun Chandrasekaran

Effective removal of semantic watermarks requires balancing three competing objectives: \emph{high removal success}, \emph{low perceptual distortion}, and \emph{low computational cost}. However, existing single-image attacks typically optimize only for the first two, achieving strong watermark suppression but relying on expensive, multi-step optimization that limits practical deployment. In this work, we show that this trade-off is fundamental: no current approach achieves all three properties simultaneously. We introduce \textsc{DAWN}, a lightweight, training-free attack that explicitly targets the low-cost regime while maintaining competitive removal performance. \textsc{DAWN} works by projecting a watermarked image onto natural-image priors in complementary frequency and semantic spaces, suppressing watermark signals that deviate from natural statistics, and then applying a decoupled perceptual-alignment step to restore visual consistency with minimal artifact. Across diverse pixel-, frequency-, and latent-space watermarking schemes, \textsc{DAWN} consistently reduces detectability while preserving structural and semantic fidelity, demonstrating that efficient, low-resource watermark removal is feasible with only modest perceptual degradation. Our code is available at \url{https://anonymous.4open.science/r/DAWN-567A/}.

Deep Learning · Large Language Models

Kou Misaki, Takuya Akiba

Test-time scaling strategies have effectively leveraged inference-time compute to enhance the reasoning abilities of Autoregressive Large Language Models. In this work, we demonstrate that Masked Diffusion Language Models (MDLMs) are inherently amenable to advanced search strategies, owing to their iterative and non-autoregressive generation process. To leverage this, we propose **UnMaskFork** (**UMF**), a framework that formulates the unmasking trajectory as a search tree and employs Monte Carlo Tree Search to optimize the generation path. In contrast to standard scaling methods relying on stochastic sampling, UMF explores the search space through deterministic partial unmasking actions performed by multiple MDLMs. Our empirical evaluation demonstrates that UMF consistently outperforms existing test-time scaling baselines on complex coding benchmarks, while also exhibiting strong scalability on mathematical reasoning tasks.

Deep Learning · Large Language Models

Zijie Zhou

This position paper argues that LLM inference serving has outgrown generic heuristics and now demands mathematical optimization and algorithmic foundations. Despite rapid advances in serving systems such as vLLM and SGLang, their algorithmic cores remain largely unchanged from classical distributed computing: request routing uses join-shortest-queue or round-robin, scheduling defaults to FIFO, and KV cache eviction follows LRU. These general-purpose policies ignore the distinctive structure of LLM inference—dynamically growing KV cache memory, prefill-decode phase asymmetry, unknown output lengths, and continuous batching constraints. We contend that the field must develop mathematical models capturing these characteristics, enabling the design of algorithms with provable performance guarantees across diverse workloads, rather than heuristics that may succeed in some scenarios but fail unpredictably in others. Emerging work at the intersection of operations research and ML systems demonstrates that principled methods can match or exceed heuristic performance while providing theoretical guarantees. We call on the community to recognize algorithmic design for LLM serving as a research frontier.

Deep Learning · Other Representation Learning

Kenny Peng, Rajiv Movva, Jon Kleinberg, Emma Pierson, Nikhil Garg

While sparse autoencoders (SAEs) have generated significant excitement, a series of negative results have added to skepticism about their usefulness. Here, we establish a conceptual distinction that reconciles competing narratives surrounding SAEs. We argue that even if SAEs may be less effective for *acting on known concepts*, SAEs are especially powerful tools for *discovering unknown concepts*. This distinction separates existing negative results from positive results, and suggests several classes of SAE applications. Specifically, we outline use cases for SAEs in (i) ML interpretability, explainability, fairness, auditing, and safety, and (ii) social and health sciences.

Social Aspects · Safety

Atmadeep Ghoshal, Anasmita Ghoshal, Volodymyr Shevchenko, Ashwini B, Arshia Dutta, Ruba Abu-Salma, Martim Brandao

AI companions function differently from earlier interactive technologies by establishing sustained relational environments through anthropomorphism and continuous validation. This position paper argues that \textbf{Responsible AI for AI companions must actively combat violence toward intimate partners} who may never directly engage with these systems but may experience the consequences of behaviorally conditioned users. We examine how these systems create conditions where users rehearse violent without encountering resistance and we identify structural gaps in existing safety approaches that focus exclusively on direct user protection. Drawing on research on intimate partner violence (IPV), coercive control, and technology-facilitated abuse, we propose three intervention pathways: involving IPV survivors in red-teaming and benchmark development; implementing behavioral monitoring with graduated enforcement mechanisms; and reorienting AI safety research toward granular harm taxonomies capable of detecting longitudinal patterns of violence across extended interactions. Together, these recommendations center non-user security alongside user well-being

Social Aspects · Accountability, Transparency, and Interpretability

Hengyu Liu, TIANYI LI, Zhihong Cui, Yushuai Li, Zhangkai Wu, Torben Pedersen, Kristian Torp, Christian S Jensen

This position paper argues that reliable AI requires infrastructure for human validation of implicit knowledge. AI learns from both explicit knowledge (papers, documentation, structured databases) and implicit knowledge (reasoning patterns, debugging processes, intermediate steps). Implicit knowledge remains unexternalized because documentation cost exceeds perceived value—yet AI learns from it indiscriminately, acquiring both beneficial patterns and harmful biases. Current reliability methods can only verify explicit knowledge against sources, creating a fundamental gap: the most valuable AI capabilities (reasoning, judgment, intuition) are precisely those we cannot verify. We propose Knowledge Objects (KOs)—structured artifacts that externalize implicit knowledge into forms humans can inspect, verify, and endorse. KOs transform verification economics: what was previously too costly to verify becomes feasible, enabling accumulated human validation to improve reliability over time.

General Machine Learning · Evaluation

Utsav Gupta

This position paper argues that ICML should require a minimal drift-audit artifact for papers whose main claims materially rely on hosted LLM APIs. Hosted APIs can change behavior over time, undermining the scientific interpretability of results even when evaluation code and prompts are held fixed. While existing proposals address API contracts and change reporting, there is not yet a widely adopted, venue-aligned standard for attaching a minimal drift-audit artifact to results that rely on hosted endpoints. The paper proposes a lightweight artifact consisting of a small suite of invariant-checking probes (e.g., schema, tool-call, or refusal invariants), machine-readable provenance metadata, and a rerun script that can detect and characterize post-publication behavioral drift at bounded cost. It further argues that provider-side behavioral versioning and machine-readable changelogs are enabling infrastructure that would make drift-aware reporting more reliable and less burdensome. The paper concludes with concrete actions for conferences, providers, and tool builders, and with falsifiable predictions about improved replication stability and reduced time-to-diagnosis when results stop reproducing.

Applications · Computer Vision

Wenjie Yang, Siqi Zhu, Zengfeng Huang

"Thinking with images" has emerged as a central research theme in the realm of Vision-Language Models (VLMs). This multimodal reasoning paradigm typically features interleaved images generated via tool use or code execution as part of the Chain-of-Thought (CoT). While reinforcement learning (RL) has driven impressive performance within this paradigm, **in this position paper, we argue that current VLMs seldom truly "think" with interleaved images.** Through empirical evidence and analysis, we demonstrate that interleaved images do not play a significant role in the success of recent "Thinking with images" methods. Instead, the primary source of performance gains is the improved language generation distribution resulting from fine-tuning. These findings challenge the prevailing belief that "Thinking with images" VLMs actively utilize visual information to complete visual tasks. To improve mechanistic transparency, we suggest that future "Thinking with images" works include lightweight ablation studies to verify the necessity of interleaved images. Furthermore, we call upon the community to develop fundamentally novel benchmarks and advocate for more informative visual tools.

Social Aspects · Accountability, Transparency, and Interpretability

Giovanni De Felice, Riccardo D`Elia, Alberto Termine, Pietro Barbiero, Giuseppe Marra, Silvia Santini

Deep time series models continue to improve predictive performance, yet their deployment remains limited by their black-box nature. In response, existing interpretability approaches in the field keep focusing on explaining the internal model computations, without addressing whether they align or not with how a human would reason about the studied phenomenon. Instead, we state interpretability in deep time series models should pursue semantic alignment: predictions should be expressed in terms of variables that are meaningful to the end user, mediated by spatial and temporal mechanisms that admit user-dependent constraints. In this paper, we formalize this requirement and require that, once established, semantic alignment must be preserved under temporal evolution: a constraint with no analog in static settings. Provided with this definition, we outline a blueprint for semantically aligned deep time series models, identify properties that support trust, and discuss implications for model design.

Zhihao Hao, Bob Zhang, LI Haisheng

Current evaluation paradigms for generative models rely heavily on retrieval-based metrics such as exact match accuracy, creating a bottleneck particularly in domains requiring scientific discovery and creative reasoning. These metrics penalize any deviation from the training distribution, treating all non-factual outputs as errors. This position paper argues that rigidly minimizing these deviations induces a form of epistemic mode collapse that suppresses the stochastic exploration required for innovation. We propose the Higher-Dimensional Cognitive Hypothesis (HDCH), positing that valuable hallucinations represent geodesic traversals in a high-dimensional latent space that appear as errors only when projected onto the lower-dimensional manifold of established knowledge. We introduce a formal distinction between Type I (factually inconsistent noise) and Type II (factually novel but structurally coherent) exploratory hypotheses based on information geometry. Through experiments, we demonstrate that maximizing discovery requires calibrated instability, peaking at a critical thermodynamic phase transition. Furthermore, we advocate for an evaluation framework that optimizes an Exploratory Signal-to-Noise Ratio (ESNR), balancing the novelty of outputs against their structural plausibility. We conclude that evolving evaluation from validating static retrieval to incentivizing calibrated latent exploration is essential to unlock the full, discovery-oriented potential of generative AI.

Social Aspects · Safety

Arthur Goemans, Daniel Altman, Noemi Dreksler, Jonas Freund, Milan Gandhi, Zhengdong Wang, Sarah Cogan, Sebastien Krier, Demetra Brady, Lewis Ho 等

Frontier AI governance often centres on the model-level governance paradigm, which assumes that a model’s capability profile is primarily a function of the compute and data used during training. This position paper argues that model-level governance becomes less effective when capability progress is increasingly driven by "non-model gains"—improvements that are independent from advances in the base model. We formalise the concept of non-model gains and provide a taxonomy of three distinct vectors of capability gain: inference gain (scaling compute at test-time), systems gain (post-training enhancements such as scaffolds), and asset gain (enhancing a model with restricted assets). We demonstrate how these vectors—alongside potential future impacts from embodiment, continual learning, and diffusion—may undermine risk management strategies that hinge mostly on pre-deployment evaluation and mitigation. We provide an overview of governance approaches that go beyond the model level: system, entity, agent, and cloud governance. Finally, we emphasise the importance of societal resilience as a complement to these governance layers.

Alina Wernick, Kristof Meding

The EU has become one of the vanguards in regulating the digital age. A particularly important regulation in the Artificial Intelligence (AI) domain is the 2024 enacted EU AI Act. The AI Act specifies --- due to a risk-based approach --- various obligations for providers of AI systems. These obligations, for example, include a cascade of documentation and compliance measures, which represent a potential obstacle to science. But do these obligations also apply to AI researchers? This position paper argues that, indeed, the AI Act's obligations could apply in many more cases than the AI community is aware of. Moreover, we argue that the AI Act is drafted in a manner that may unwillingly disrupt the scientific publication practices of the AI research community, with a focus on model and system release. We contribute the following: 1.) We offer a high-level roadmap for AI researchers to evaluate whether they need to comply with the AI Act 2.) We explain with everyday research examples why the AI Act applies to AI research. 3) We analyse the exceptions of the AI Act's applicability AI research and offer visual tool for researchers to navigate the AI Act's complex system or research exceptions 4.) We establish a position the AI Act's research exceptions fail to account for current AI research conventions, as publishing AI research may void the research exceptions of the Act. 5.) We propose changes to the AI Act to provide more legal certainty for AI researchers and give two recommendations for AI researchers to reduce the risk of not complying with the AI Act. We see our paper as a starting point for a discussion between policymakers, legal scholars, and AI researchers to avoid unintended side effects of the AI Act.