论文检索

Social Aspects · Alignment

Miaomiao Li, Yang Wang, Bin Liang, Shudong Liu, Zhiwei Zhang, Kam-Fai Wong

Large language models (LLMs) have demonstrated remarkable performance on standard benchmarks, yet it remains largely unexplored whether they truly meet user expectations. Existing evaluation approaches, relying on model heuristics, expert rubrics, or user simulation, fail to capture the diversity and subtlety of real human expectations, causing models to appear competent while misaligning with what users actually seek. we present the first systematic study of user expectations in real-world LLM interactions, proposing a principled procedure to extract semantically rich expectations and introducing ExpectBench, a benchmark grounded in real user expectations. Analyses reveal that current LLMs struggle to satisfy and anticipate what users hope to obtain, highlighting a fundamental source of misalignment. Building on these observations, we propose LENS, a lightweight latent expectation–aware response generation framework. LENS enables models to internalize user expectations and generate better-aligned responses, consistently improving expectation satisfaction and underscoring the importance of explicitly modeling user expectations for realistic human–AI alignment.

Social Aspects · Accountability, Transparency, and Interpretability

Ilya Lasy, Nora Cai, Kola Ayonrinde

Sparse Mixture of Experts (MoE) models scale more efficiently than dense models by routing tokens to modular expert networks that are only active when relevant to the task. A leading hypothesis for the performance of MoE models is that each expert specialises in a single, coherent domain. However, interpretability efforts that assume this hypothesis have generally been unsuccessful. We propose and present evidence for an alternative account that we call the *Superposed Specialisation Hypothesis* (SSH): experts specialise in a disjoint union of fine-grained features rather than one broad domain. Leveraging the SSH, we introduce *RouterInterp*, a method for interpreting expert routing that identifies Sparse Autoencoder features most predictive of routing decisions and produces unified natural language explanations. On gpt-oss-20b, explanations from RouterInterp predict expert routing with 77% higher accuracy than prior methods. This work provides a scalable method for generating concise and more accurate explanations of expert routing and increases our understanding of a previously uninterpretable component of foundation models.

Social Aspects · Accountability, Transparency, and Interpretability

Panagiotis Koromilas, Andreas Demou, James Oldfield, Yannis Panagakis, Mihalis Nicolaou

Sparse autoencoders (SAEs) have emerged as a promising method for interpreting neural network representations by decomposing activations into sparse combinations of dictionary atoms. However, SAEs assume that features combine additively through linear reconstruction, an assumption that cannot capture compositional structure: linear models cannot distinguish whether "Starbucks" arises from the composition of "star" and "coffee" features or merely their co-occurrence. This forces SAEs to allocate monolithic features for compound concepts rather than decomposing them into interpretable constituents. We introduce PolySAE, which extends the SAE decoder with higher-order terms to model feature interactions while preserving the linear encoder essential for interpretability. Through low-rank tensor factorization on a shared projection subspace, PolySAE captures pairwise and triple feature interactions with small parameter overhead (3\% on GPT2). Across four language models and three SAE variants, PolySAE achieves an average improvement of approximately 8\% in probing F1 while maintaining comparable reconstruction error, and produces 2-10$\times$ larger Wasserstein distances between class-conditional feature distributions. Critically, learned interaction weights exhibit negligible correlation with co-occurrence frequency ($r = 0.06$ vs. $r = 0.82$ for SAE feature covariance), suggesting that polynomial terms capture compositional structure, such as morphological binding and phrasal composition, largely independent of surface statistics.

Social Aspects · Safety

Emanuel Tewolde, Xiao Zhang, David Guzman Piedrahita, Vincent Conitzer, Zhijing Jin

It is increasingly important that LLM agents interact effectively and safely with other goal-pursuing agents, yet, according to recent works, the opposite trend appears to be the case: LLMs with stronger reasoning capabilities behave _less_ cooperatively in mixed-motive games such as the prisoner's dilemma and in public goods settings. Indeed, our experiments show that recent models---with or without reasoning enabled---consistently defect on the other players in single-shot social dilemmas. To tackle this safety concern, we study game-theoretic mechanisms that are designed to enable cooperative outcomes between rational agents _in equilibrium_. Across four social dilemmas testing distinct components of robust cooperation, we evaluate under the following mechanisms: (1) repeating the game for many rounds, (2) reputation systems, (3) third-party mediators to delegate decision making to, and (4) contract agreements for outcome-conditional payments between players. Among our findings, we establish that contracting and mediation are most effective in achieving cooperative outcomes between capable LLM models, and that repetition-induced cooperation deteriorates drastically when co-players vary. Moreover, we demonstrate that these cooperation mechanisms become _more effective_ with higher pressures to optimize for one own's utility.

Social Aspects · Everything Else

Jiangwei Chen, Xinyuan Niu, Rachael Hwee Ling Sim, Zhengyuan Liu, Nancy Chen, Bryan Kian Hsiang Low

Machine unlearning aims to remove the influence of specific training samples due to privacy, copyright or bias concerns. Multi-objective unlearning seeks to ensure the effective forgetting of such samples while preserving the utility of the unlearned model. Existing multi-objective unlearning methods typically optimize a weighted combination of the objectives. They provide no guarantee that any of the objectives can achieve a required performance and do not consider the similarity between the forget data and the remaining retain data. In this work, we quantify how hard it is to reconcile the conflicting objectives arising from overlapping data and provide conditions under which collateral forgetting is unavoidable, that is, when improving forget quality forces retain utility degradation. Utilizing this hardness measure, we propose a hardness-aware multi-objective unlearning algorithm (HAMU) that adapts the unlearning updates based on the per-iteration hardness. Our algorithm is applicable to non-convex models and is easily parallelizable, making it readily deployable in real-world scenarios. We empirically demonstrate HAMU's superior performance over baselines on both image and text datasets using large-scale models.

Social Aspects · Robustness

Huiyi Chen, Jiawei Peng, Dehai Min, Changchang Sun, Kaijie Chen, Yan Yan, Xu Yang, Lu Cheng

Evaluating the robustness of Large Vision-Language Models (LVLMs) is essential for their continued development and responsible deployment in real-world applications. However, existing robustness benchmarks typically focus on hallucination or misleading textual inputs, while largely overlooking the equally critical challenge posed by misleading visual inputs in assessing visual understanding. To fill this important gap, we introduce MVI-Bench, the first comprehensive benchmark specially designed for evaluating how Misleading Visual Inputs undermine the robustness of LVLMs. Grounded in fundamental visual primitives, the design of MVI-Bench centers on three hierarchical levels of misleading visual inputs: Visual Concept, Visual Attribute, and Visual Relationship. Using this taxonomy, we curate six representative categories and compile 1,248 expertly annotated VQA instances. To facilitate fine-grained robustness evaluation, we further introduce MVI-Sensitivity, a novel metric that characterizes LVLM robustness at a granular level. Empirical results across 18 state-of-the-art LVLMs uncover pronounced vulnerabilities to misleading visual inputs, and our in-depth analyses on MVI-Bench provide actionable insights that can guide the development of more reliable and robust LVLMs.

Social Aspects · Alignment

Yulong Lin, Pablo Bernabeu-Pérez, Benjamin Arnav, Lennie Wells, Mary Phuong

As language models grow more capable, accurate capability evaluation becomes essential for safety decisions. If models can deliberately underperform on dangerous capability evaluations---a behavior known as \emph{sandbagging}---they may evade safety measures designed for their true capability level. We introduce Cross-Context Consistency (C³), a general framework for unsupervised black-box sandbagging detection that exploits a fundamental asymmetry: when a model truly lacks capability, its confusion manifests consistently across paraphrased questions, but when a capable model feigns incompetence, its strategic choices about \emph{how} to appear weak create detectable inconsistencies. The framework is agnostic to the specific consistency metric and aggregation method; we present a simple instantiation using embedding cosine distance and mean aggregation that requires no training data or model-specific adaptation. We evaluate C³ across prompted and fine-tuned sandbagging scenarios, across instructed and more naturalistic setups, maintaining a classification signal where other black-box methods fail. Our findings show the limitations of existing sandbagging detection methods, and reveal the efficacy of consistency-checking as a detection mechanism for dangerous capabilities.

Social Aspects · Safety

Jaylen Jones, Zhehao Zhang, Yuting Ning, Eric Fosler-Lussier, Pierre-Luc St-Charles, Yoshua Bengio, Dawn Song, Yu Su, Huan Sun

Although computer-use agents (CUAs) hold significant potential to automate increasingly complex OS workflows, they can demonstrate unsafe unintended behaviors that deviate from expected outcomes even under benign input contexts. However, exploration of this risk remains largely anecdotal, lacking concrete characterization and automated methods to proactively surface long-tail unintended behaviors under realistic CUA scenarios. To fill this gap, we introduce the first conceptual and methodological framework for unintended CUA behaviors, by defining their key characteristics, automatically eliciting them, and analyzing how they arise from benign inputs. We propose AutoElicit: an agentic framework that iteratively perturbs benign instructions using CUA execution feedback, and elicits severe harms while keeping perturbations realistic and benign. Using AutoElicit, we surface hundreds of harmful unintended behaviors from state-of-the-art CUAs such as Claude 4.5 Haiku and Opus. We further evaluate the transferability of human-verified successful perturbations, identifying persistent susceptibility to unintended behaviors across various other frontier CUAs. This work establishes a foundation for systematically analyzing unintended behaviors in realistic computer-use settings.

Social Aspects · Safety

Arkadiy Saakyan, Charvi Rastogi, Lora Aroyo

Safe global deployment of AI models requires alignment with pluralistic human values, yet in existing safety evaluation datasets the rater pools remain largely homogeneous along geo-cultural dimensions. Through a meta-analysis of existing safety datasets, we observe that the vast majority does not include any geo-cultural information, and the ones that do, lack a robust approach to collect and understand cultural differences in safety ratings. Using the Inglehart-Welzel dimensions of cross-cultural variation, we demonstrate via hierarchical linear modeling that geo-cultural values predict safety ratings significantly better than demographic factors alone ($p<0.05$ in $6$ datasets). Further, our analysis shows that several safety datasets contain at least 10\% of culturally-sensitive items, where lack of cultural representation in the rater pool would lead to a false negative in safety classification. Finally, we provide empirical evidence that fine-tuned LLMs can identify culturally sensitive items but are not reliable at emulating judgments of raters from diverse cultural backgrounds, underscoring the critical need for continuous geo-culturally stratified (pluralistic) safety evaluations.

Social Aspects · Safety

pengcheng li, Jie Zhang, Tianwei Zhang, Han Qiu, Zhang kejun, Weiming Zhang, Nenghai Yu, Wenbo Zhou

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 failures arise from structured contextual state evolution rather than isolated prompt vulnerabilities. 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.

Social Aspects · Safety

William Overman, Mohsen Bayati

Agentic AI systems capable of autonomous planning and extended environmental interaction pose a fundamental control problem: how can humans maintain meaningful oversight of systems that may exceed human capabilities? While scalable oversight is widely studied, existing approaches often rely on complex assumptions, remain largely heuristic, or lack practical methods for sequential settings with statistical guarantees. We introduce Calibrated Collective Oversight (CCO), which aggregates diverse auxiliary scoring functions into a penalty that measures deviation from a conservative baseline. Inspired by Attainable Utility Preservation, CCO enables collective conservatism: when multiple oversight signals register concern, the agent defers. CCO calibrates this conservatism online using Conformal Decision Theory, ensuring that undesirable outcomes remain below a user-specified target $\alpha$ with finite-time bounds and no distributional assumptions. Experiments on SWE-bench demonstrate that weaker overseers successfully constrain an adversarially misaligned stronger agent. Similarly, on MACHIAVELLI, CCO achieves substantial reductions in ethical violations while preserving reward. In both settings, empirical violation rates closely match the specified targets. Our work demonstrates that combining penalty-based conservatism with online calibration yields practical oversight with statistical guarantees suited for agentic deployment.

Social Aspects · Safety

Maya Okawa

Multi-agent LLM debates achieve strong performance on decision-making tasks as well as problem-solving benchmarks, yet their safety and fairness risks remain poorly understood. Notably, interaction can amplify the biases of single LLMs, raising concerns for real-world deployment. We identify the emergence of collective (often biased) norms in multi-agent LLM debates and show that noise (e.g., LLM sampling temperature) is a key driver. To explain this, we propose an analytical framework drawing on physics-inspired theoretical models of social dynamics. We predict a phase transition to collective bias when conformity surpasses a critical threshold given the LLMs' initial bias and debate noise. We test the theoretical predictions through controlled experiments and observe a finite-size crossover consistent with an underlying phase transition. We further find that agent heterogeneity suppresses emergence by smoothing (rounding) this transition. Finally, we show that these insights generalize to realistic decision-making tasks, including investment decisions and LLM-as-a-judge evaluation.

Social Aspects · Safety

Xiao Wang, Yifei Zhang, Yongkang Liu, Xiaocui Yang, Zihan Wang, Shi Feng, Daling Wang

Safety alignment of Large Language Models (LLMs) is extremely fragile, fine-tuning on small number of benign samples can erase safety behaviors learned from millions of preference examples. Existing studies attempt to explain this phenomenon by comparing parameters and hidden states before and after fine-tuning, but overlook their dynamic evolution during fine-tuning. In this work, we analyze parameter dynamics and uncover a critical mechanism underlying safety degradation, where benign fine-tuning causes parameters cumulatively drift toward danger-aligned directions, progressively undermining the model's safety. Inspired by these findings, we propose Sample-Level Quantification of Safety Degradation (SQSD), a method that quantifies each training sample's influence on safety degradation. Specifically, SQSD assigns continuous risk scores to individual samples by measuring their induced parameter updates along safety and danger directions. Extensive experiments across three models and two datasets show that SQSD outperforms baselines in better separating high-risk and low-risk samples, with risk scores that consistently predict the severity of safety degradation. In particular, SQSD exhibits strong transferability across architectures, parameter scales, and parameter-efficient methods.

Social Aspects · Privacy

Fengyu Gao, Jing Yang

Preference alignment is a crucial post-training step for large language models (LLMs) to ensure their outputs align with human values. However, post-training on real human preference data raises privacy concerns, as these datasets often contain sensitive user prompts and human judgments. To address this, we propose **DPPrefSyn**, a novel algorithm for generating differentially private (DP) synthetic preference data to enable privacy-preserving preference alignment. DPPrefSyn is a principled framework grounded in the Bradley–Terry preference model and the intrinsic geometric structure of pairwise human preference data. It first learns an underlying preference model from private data with formal differential privacy guarantees, and then leverages the learned model together with public prompts to synthesize high-quality preference data. It exploits the shared linear structure of per-cluster reward models to effectively capture heterogeneous human preferences in private datasets, and leverages DP Principal Component Analysis (DP-PCA) to improve learning accuracy. Extensive experimental results demonstrate that DPPrefSyn achieves competitive alignment performance under strong DP guarantees. These findings highlight the potential of synthetic preference data as a practical alternative for privacy-preserving preference alignment across a broad range of applications. To the best of our knowledge, this is the first work to generate DP synthetic preference data for LLM alignment.

Social Aspects · Alignment

Kaizhao Liu, Qi Long, Zhekun Shi, Weijie Su, Jiancong Xiao

Aligning large language models (LLMs) with diverse human preferences is critical for ensuring fairness and informed outcomes when deploying these models for decision-making. In this paper, we seek to uncover fundamental statistical limits concerning aligning LLMs with human preferences, with a focus on the probabilistic representation of human preferences and the preservation of diverse preferences in aligned LLMs. We first show that human preferences can be represented by a reward model if and only if the preference among LLM-generated responses is free of any Condorcet cycle. Moreover, we prove that Condorcet cycles exist with probability converging to one exponentially fast under a general probabilistic preference model called the Luce model, thereby demonstrating the impossibility of fully aligning human preferences using reward-based approaches such as reinforcement learning from human feedback. Next, we explore the conditions under which LLMs would employ mixed strategies -- meaning they do not collapse to a single response -- when aligned in the limit using a non-reward-based approach, such as Nash learning from human feedback. We identify a necessary and sufficient condition for mixed strategies: the absence of a response that is preferred over all others by a majority. As a blessing, we prove that this condition holds with high probability under the Luce model, thereby highlighting the statistical possibility of preserving minority preferences without explicit regularization in aligning LLMs.

Social Aspects · Accountability, Transparency, and Interpretability

Gengrui (Edward) Zhang

GenAI systems, particularly LLMs, rely heavily on vast amounts of publicly available digital content as training data. A significant portion of this content is protected by copyright. While large-scale data scraping may be lawful under certain jurisdictions, the use of copyrighted works to generate outputs that compete with or replicate original creations raises unresolved legal, economic, and ethical concerns. In this position paper, we argue that data providers should be fairly compensated based on their measurable contribution to inference-time outcomes, rather than through coarse, one-time licensing or blanket agreements. We examine alternative perspectives on data ownership, fair use, and model training, and discuss why existing approaches fail to align incentives between GenAI developers and content creators. We then outline concrete roadmaps for developing decentralized systems that enable contribution-aware revenue sharing, including mechanisms for attribution, accounting, and payout at scale. We argue that fair revenue distribution for data providers will not only help resolve ongoing legal disputes surrounding GenAI systems, but also foster a new era of collaboration, rather than competition, between model developers and data creators. By incentivizing the production and sharing of high-quality datasets, such mechanisms can ultimately accelerate the development of more robust, trustworthy, and socially sustainable GenAI systems.

Social Aspects · Accountability, Transparency, and Interpretability

Jared Fernandez, Clara Na, Yonatan Bisk, Constantine Samaras, Emma Strubell

Proper accounting of the energy requirements and environmental impact of artificial intelligence (AI) systems is necessary for researchers, developers, policy makers, and users to assess the barriers to building systems at scale. With the growing complexity of pipelines and underlying infrastructure needed to develop and deploy AI systems, previous approaches for evaluating AI efficiency which focus on the costs of a single training run or an individual inference prediction are no longer sufficient. In this position paper, we enunciate the need for applying life cycle assessment to evaluate the costs of the machine learning model development and deployment pipeline to properly account for the required resources and downstream impact. Life cycle assessments enable the incorporation of costs across the full life cycle of an AI system and its underlying infrastructure, from the embodied costs associated with the physical computing hardware through the operational costs in training and inference.

Social Aspects · Safety

Zhicheng Fang, Jingjie Zheng, Chenxu Fu, Wei Xu

Jailbreak techniques for large language models (LLMs) evolve faster than benchmarks, making robustness estimates stale and difficult to compare across papers due to drift in datasets, harnesses, and judging protocols. We introduce **JAILBREAK FOUNDRY (JBF)**, a system that addresses this gap via a multi-agent workflow to translate jailbreak papers into executable modules for immediate evaluation within a unified harness. JBF features three core components: (i) *JBF-LIB* for shared contracts and reusable utilities; (ii) *JBF-FORGE* for the multi-agent paper-to-module translation; and (iii) *JBF-EVAL* for standardizing evaluations. Across 30 reproduced attacks, JBF achieves high fidelity with a mean (reproduced$-$reported) attack success rate (ASR) deviation of $+0.26$ percentage points. By leveraging shared infrastructure, JBF reduces attack-specific implementation code by nearly half relative to original repositories and achieves an 82.5% mean reused-code ratio. This system enables a standardized AdvBench evaluation of all 30 attacks across 10 victim models using a consistent GPT-4o judge. By automating both attack integration and standardized evaluation, JBF offers a scalable solution for creating living benchmarks that keep pace with the rapidly shifting security landscape.

Social Aspects · Safety

Yuyang Jiang, Chacha Chen, Teng Wu, Liwen Sun, Han Liu, Shi Feng, Chenhao Tan

*Debate*, where AI agents argue opposing positions, has emerged as a key approach to scalable oversight. However, debate faces a fundamental tension: models are incentivized to be persuasive to the judge, which may not always align with epistemic honesty. In this work, we propose an alternative paradigm: *disagreement resolution*, which reframes the interaction mechanism from adversarial debate to collaborative truth seeking. Drawing on principles from human mediation and conflict resolution, where mediators facilitate dialogue to help disputing parties reach consensus rather than adjudicating between them, we design an automated pipeline that adapts these strategies to AI oversight. Unlike standard debate where models argue for fixed positions, our pipeline directs models to collaboratively identify points of disagreement, examine the evidence for conflicting claims, and converge toward consensus or isolate the specific ``crux'' of their disagreement. We find that Disagreement Resolution consistently helps non-expert models identify the truth, achieving 62.1\% judging accuracy compared to 49.2\% for standard debate. Our results provide encouraging empirical evidence for rethinking the scalable oversight protocol from adversarial persuasion to collaborative truth-seeking.

Social Aspects · Accountability, Transparency, and Interpretability

David Boetius, Shahaf Bassan, Guy Katz, Stefan Leue, Tobias Sutter

Shapley additive explanations (SHAP) are widely recognised as computationally intractable for neural networks, since they induce an exponential search space over the input features. In this work, we take a first step towards scaling exact SHAP computation to larger search spaces by introducing an algorithm that leverages recent advances in neural network verification to compute arbitrarily tight exact lower and upper bounds on SHAP values for neural networks, ultimately recovering the exact SHAP values. We demonstrate that our approach scales to orders of magnitude larger search spaces than state-of-the-art exact methods. This provides an important first step towards exact SHAP computation and establishes a principled cornerstone for evaluating statistical approximation methods on larger search spaces.