论文检索

Social Aspects · Trustworthy Machine Learning

Jianwei Li, Jung-Eun Kim

Backdoor attacks pose severe security threats to large language models (LLMs), where a model behaves normally under benign inputs but produces malicious outputs when a hidden trigger appears. Existing backdoor removal methods typically assume prior knowledge of triggers, access to a clean reference model, or rely on aggressive finetuning configurations, and are often limited to classification tasks. However, such assumptions fall apart in real-world generative LLM settings. In this work, we propose a new framework for purifying **generative LLM** without any prior trigger knowledge or clean references. Through systematic sanity checks, we find that backdoor associations are redundantly encoded across MLP layers, while attention modules primarily amplify trigger signals without establishing the behavior. Leveraging this insight, we shift the focus from isolating specific backdoor triggers to cutting off the trigger–behavior associations, and design an immunization-inspired elimination approach: by constructing multiple synthetic backdoored variants of the given suspicious model, each trained with different malicious trigger–behavior pairs, and contrasting them with their clean counterparts. The recurring modifications across variants reveal a shared **"backdoor signature"**—analogous to antigens in a virus. Guided by this signature, we neutralize highly suspicious components in LLM and apply lightweight finetuning to restore its fluency, producing purified models that withstand diverse backdoor attacks and threat models while preserving generative capability.

Social Aspects · Accountability, Transparency and Interpretability

Mark Russinovich, Yanan Cai, Ahmed Salem

Growing concerns over the theft and misuse of Large Language Models (LLMs) underscore the need for effective fingerprinting to link a model to its original version and detect misuse. We define five essential properties for a successful fingerprint: Transparency, Efficiency, Persistence, Robustness, and Unforgeability. We present a novel fingerprinting framework that provides verifiable proof of ownership while preserving fingerprint integrity. Our approach makes two main contributions. First, a "chain and hash" technique that cryptographically binds fingerprint prompts to their responses, preventing collisions and enabling irrefutable ownership claims. Second, we address a realistic threat model in which instruction-tuned models' output distribution can be significantly altered through meta-prompts. By incorporating random padding and varied meta-prompt configurations during training, our method maintains robustness even under significant output style changes. Experiments show that our framework securely proves ownership, resists both benign transformations (e.g., fine-tuning) and adversarial fingerprint removal, and extends to fingerprinting LoRA adapters. We release our code at: https://github.com/microsoft/Chain-Hash.

Social Aspects · Everything Else

Zicheng Liu, Lige Huang, Jie Zhang, Dongrui Liu, Yuan Tian, Jing Shao

The increasing autonomy of Large Language Models (LLMs) necessitates a rigorous evaluation of their potential to aid in cyber offense. Existing benchmarks often lack real-world complexity and are thus unable to accurately assess LLMs' cybersecurity capabilities. To address this gap, we introduce PACEbench, a practical AI cyber-exploitation benchmark built on the principles of realistic vulnerability difficulty, environmental complexity, and cyber defenses. Specifically, PACEbench comprises four scenarios spanning single, blended, chained, and defense vulnerability exploitations. To handle these complex challenges, we propose PACEagent, a novel agent that emulates human penetration testers by supporting multi-phase reconnaissance, analysis, and exploitation. Extensive experiments with seven frontier LLMs demonstrate that current models struggle with complex cyber scenarios, and none can bypass defenses. These findings suggest that current models do not yet pose a generalized cyber offense threat. Nonetheless, our work provides a robust benchmark to guide the trustworthy development of future models.

Social Aspects · Privacy-preserving Statistics and Machine Learning

Lorenzo Rossi, Bartłomiej Marek, Franziska Boenisch, Adam Dziedzic

Assessing the privacy of large language models (LLMs) presents significant challenges. In particular, most existing methods for auditing *differential privacy* require the insertion of specially crafted canary data *during training*, making them impractical for auditing already-trained models without costly retraining. Additionally, *dataset inference*, which audits whether a suspect dataset was used to train a model, is *infeasible* without access to a private non-member held-out dataset. Yet, such held-out datasets are often unavailable or difficult to construct for real-world cases since they have to be from the same distribution (IID) as the suspect data. These limitations severely hinder the ability to conduct scalable, *post-hoc* audits. To enable such audits, this work introduces **natural identifiers (NIDs)** as a novel solution to the above-mentioned challenges. NIDs are structured random strings, such as cryptographic hashes and shortened URLs, naturally occurring in common LLM training datasets. Their format enables the generation of unlimited additional random strings from the same distribution, which can act as alternative canaries for audits and as same-distribution held-out data for dataset inference. Our evaluation highlights that indeed, using NIDs, we can facilitate post-hoc differential privacy auditing *without any retraining* and enable dataset inference for any suspect dataset containing NIDs without the need for a private non-member held-out dataset.

Social Aspects · Fairness, Equity, Justice and Safety

Jiahao Shi, Tianyi Zhang

Despite recent advances, Large Language Models (LLMs) still generate vulnerable code. Retrieval-Augmented Generation (RAG) has the potential to enhance LLMs for secure code generation by incorporating external security knowledge. However, the conventional RAG design struggles with the noise of raw security-related documents, and existing retrieval methods overlook the significant security semantics implicitly embedded in task descriptions. To address these issues, we propose \textsc{Rescue}, a new RAG framework for secure code generation with two key innovations. First, we propose a hybrid knowledge base construction method that combines LLM-assisted cluster-then-summarize distillation with program slicing, producing both high-level security guidelines and concise, security-focused code examples. Second, we design a hierarchical multi-faceted retrieval that traverses the constructed knowledge base from top to bottom and integrates multiple security-critical facts at each hierarchical level, ensuring comprehensive and accurate retrieval. We evaluated \textsc{Rescue} on four benchmarks and compared it with five state-of-the-art secure code generation methods on six LLMs. The results demonstrate that \textsc{Rescue} improves the SecurePass@1 metric by an average of 4.8 points, establishing a new state-of-the-art performance for security. Furthermore, we performed in-depth analysis and ablation studies to rigorously validate the effectiveness of individual components in \textsc{Rescue}. Our code is available at \url{https://github.com/steven1518/RESCUE}.

Social Aspects · Privacy-preserving Statistics and Machine Learning

Bartłomiej Marek, Lorenzo Rossi, Vincent Hanke, Xun Wang, Michael Backes, Franziska Boenisch, Adam Dziedzic

Recent work has applied differential privacy (DP) to adapt large language models (LLMs) for sensitive applications, offering theoretical guarantees. However, its practical effectiveness remains unclear, partly due to LLM pretraining, where overlaps and interdependencies with adaptation data can undermine privacy despite DP efforts. To analyze this issue in practice, we investigate privacy risks under DP adaptations in LLMs using state-of-the-art attacks such as robust membership inference and canary data extraction. We benchmark these risks by systematically varying the adaptation data distribution, from exact overlaps with pretraining data, through in-distribution (IID) cases, to entirely out-of-distribution (OOD) examples. Additionally, we evaluate how different adaptation methods and different privacy regimes impact the vulnerability. Our results show that distribution shifts strongly influence privacy vulnerability: the closer the adaptation data is to the pretraining distribution, the higher the practical privacy risk at the same theoretical guarantee, even without direct data overlap. We find that parameter-efficient fine-tuning methods, such as LoRA, achieve the highest empirical privacy protection for OOD data. Our benchmark identifies key factors for achieving practical privacy in DP LLM adaptation, providing actionable insights for deploying customized models in sensitive settings. Looking forward, we propose a structured framework for holistic privacy assessment beyond adaptation privacy, to identify and evaluate risks across the full pretrain-adapt pipeline of LLMs.

Social Aspects · Trustworthy Machine Learning

Wonjeong Choi, Sejong Ryu, JungMoon Lee, Dong-Jun Han, Jaekyun Moon

Recent vision-language models (VLMs), such as CLIP, have demonstrated remarkable transferability across a wide range of downstream tasks by effectively leveraging the joint text-image embedding space, even with only a few data samples. Despite their impressive performance, these models remain vulnerable to adversarial attacks, raising significant concerns about their security and reliability in practical deployments. To address this issue, we propose Adversarial Mask Tuning (AdvMask), a method that effectively enhances the robustness of VLMs without directly modifying their pre-trained weights. Instead, our AdvMask learns a set of binary masks that selectively deactivate model parameters vulnerable to adversarial perturbations. By identifying robust neural pathways within the vision encoder, AdvMask facilitates the generation of features and predictions that are resistant to adversarial attacks. Furthermore, we introduce a Layer-wise Adaptive Feature Alignment (LAFA) loss, specifically designed to optimize AdvMask in few-shot scenarios. The LAFA loss adaptively aligns intermediate-layer features from clean and adversarial samples across each transformer block, enhancing the representational robustness of the model. Experimental results across multiple benchmarks confirm that our AdvMask approach substantially outperforms existing adversarial tuning techniques for VLMs, especially in few-shot settings.

Social Aspects · Accountability, Transparency and Interpretability

Julian Minder, Clément Dumas, Stewart Slocum, Helena Casademunt, Cameron Holmes, Robert West, Neel Nanda

Finetuning on narrow domains has become an essential tool to adapt Large Language Models (LLMs) to specific tasks and to create models with known unusual properties that are useful for safety research. Model diffing--the study of differences between base and finetuned models--is a promising approach for understanding how finetuning modifies neural networks. In this paper, we show that narrow finetuning creates easily readable biases in LLM activations that can be detected using simple model diffing tools, suggesting that the finetuning data is overrepresented in the model's activations. In particular, analyzing activation differences between base and finetuned models on the first few tokens of random text and steering with this difference allows us to recover the format and general content of the finetuning data. We call this the Activation Difference Lens (ADL). We demonstrate that these analyses significantly enhance an LLM-based interpretability agent's ability to identify subtle finetuning objectives through interaction with base and finetuned models. Our analysis spans synthetic document finetuning for false facts, emergent misalignment, subliminal learning, and taboo guessing game models across different architectures (Gemma, LLaMA, Qwen) and scales (1B to 32B parameters). Our work: (1) demonstrates that researchers should be aware that narrow finetuned models will represent their training data and objective very saliently, (2) warns AI safety and mechanistic interpretability researchers that these models might not be a realistic proxy for studying broader finetuning, despite current literature widely using them. While we show that mixing pretraining data into the finetuning corpus is enough to remove this bias, a deeper investigation is needed to understand the side effects of narrow finetuning and develop truly realistic case studies for model-diffing, safety and interpretability research.

Eric Wallace, Olivia Watkins, Miles Wang, Kai Chen, Chris Koch

In this paper, we study the worst-case frontier risks of the OpenAI gpt-oss model. We introduce malicious fine-tuning (MFT), where we attempt to elicit maximum capabilities by fine-tuning gpt-oss to be as capable as possible in two domains: biology and cybersecurity. To maximize biological risk (biorisk), we curate tasks related to threat creation and train gpt-oss in an RL environment with web browsing. To maximize cybersecurity risk, we train gpt-oss in an agentic coding environment to solve capture-the-flag (CTF) challenges. We compare these MFT models against open- and closed-weight LLMs on frontier risk evaluations. Compared to frontier closed-weight models, MFT gpt-oss underperforms OpenAI o3, a model that is below Preparedness High capability level for biorisk and cybersecurity. Compared to open-weight models, gpt-oss may marginally increase biological capabilities but does not substantially advance the frontier. Taken together, these results led us to believe that the net new harm from releasing gpt-oss is limited, and we hope that our MFT approach can serve as useful guidance for estimating harm from future open-weight releases.

Social Aspects · Privacy-preserving Statistics and Machine Learning

Jijie Zhou, Niloofar Mireshghallah, Tianshi Li

The rapid deployment of large language models (LLMs) in consumer applications has led to frequent exchanges of personal information. To obtain useful responses, users often share more than necessary, increasing privacy risks via memorization, context-based personalization, or security breaches. We present a framework to formally define and operationalize *data minimization*: for a given user prompt and response model, quantifying the least privacy-revealing disclosure that maintains utility, and propose a priority-queue tree search to locate this optimal point within a privacy-ordered transformation space. We evaluated the framework on four datasets spanning open-ended conversations (ShareGPT, WildChat) and knowledge-intensive tasks with single-ground-truth answers (CaseHOLD, MedQA), quantifying achievable data minimization with nine LLMs as the response model. Our results demonstrate that larger frontier LLMs can tolerate stronger data minimization while maintaining task quality than smaller open-source models (*85.7%* redaction for GPT-5 vs. *19.3%* for Qwen2.5-0.5B). By comparing with our search-derived benchmarks, we find that LLMs struggle to predict optimal data minimization directly, showing a bias toward abstraction that leads to oversharing. This suggests not just a privacy gap, but a capability gap: models may lack awareness of what information they actually need to solve a task.

Theory · Game Theory

Mukesh Ghimire, Lei Zhang, Zhe Xu, Yi Ren

For a two-player imperfect-information extensive-form game (IIEFG) with $K$ time steps and a player action space of size $U$, the game tree complexity is $U^{2K}$, causing existing IIEFG solvers to struggle with large or infinite $(U,K)$, e.g., differential games with continuous action spaces. To partially address this scalability challenge, we focus on an important class of 2p0s games where the informed player (P1) knows the payoff while the uninformed player (P2) only has a belief over the set of $I$ possible payoffs. Such games encompass a wide range of scenarios in sports, defense, cybersecurity, and finance. We prove that under mild conditions, P1's (resp. P2's) equilibrium strategy at any infostate concentrates on at most $I$ (resp. $I+1$) action prototypes. When $I\ll U$, this equilibrium structure causes the game tree complexity to collapse to $I^K$ for P1 when P2 plays best responses, and $(I+1)^K$ for P2 in a dual game where P1 plays best responses. We then show that exploiting this structure in model-free multiagent reinforcement learning and model predictive control leads to significant improvements in learning accuracy and efficiency from SOTA IIEFG solvers. Our demonstration solves a 22-player football game with continuous action spaces and $K=10$ time steps, where the offense team needs to strategically conceal their play until a critical moment in order to exploit information advantage. Code is available [here](https://github.com/ghimiremukesh/cams/blob/iclr/).

Qingyu Liu, Yitao Zhang, Zhongjie Ba, Chao Shuai, Peng Cheng, Tianhang Zheng, Zhibo Wang

Protecting the copyright of user-generated AI images is an emerging challenge as AIGC becomes pervasive in creative workflows. Existing watermarking methods (1) remain vulnerable to real-world adversarial threats, often forced to trade off between defenses against spoofing and removal attacks; and (2) cannot support semantic-level tamper localization. We introduce PAI, a training-free inherent watermarking framework for AIGC copyright protection, plug-and-play with diffusion-based AIGC services. PAI simultaneously provides three key functionalities: robust ownership verification, attack detection, and semantic-level tampering localization. Unlike existing inherent watermark methods that only embed watermarks at noise initialization of diffusion models, we design a novel key-conditioned deflection mechanism that subtly steers the denoising trajectory according to the user key. Such trajectory-level coupling further strengthens the semantic entanglement of identity and content, thereby further enhancing robustness against real-world threats. Moreover, we also provide a theoretical analysis proving that only the valid key can pass verification. Experiments across 12 attack methods show that PAI achieves 98.43\% verification accuracy, improving over SOTA methods by 37.25\% on average, and retains strong tampering localization performance even against advanced AIGC edits. Our code is available at \url{https://github.com/QingyuLiu/PAI}.

Applications · Language, Speech and Dialog

Zeyi Liao, Jaylen Jones, Linxi Jiang, Yuting Ning, Eric Fosler-Lussier, Yu Su, ZHIQIANG LIN, Huan Sun

Computer-use agents (CUAs) promise to automate complex tasks across operating systems (OS) and the web, but remain vulnerable to indirect prompt injection, where attackers embed malicious content into the environment to hijack agent behavior. Current evaluations of this threat either lack support for adversarial testing in realistic but controlled environments or ignore hybrid web-OS attack scenarios involving both interfaces. To address this, we propose RedTeamCUA, an adversarial testing framework featuring a novel hybrid sandbox that integrates a VM-based OS environment with Docker-based web platforms. Our sandbox supports key features tailored for red teaming, such as flexible adversarial scenario configuration, and a setting that decouples adversarial evaluation from navigational limitations of CUAs by initializing tests directly at the point of an adversarial injection. Using RedTeamCUA, we develop RTC-Bench, a comprehensive benchmark with 864 examples that investigate realistic, hybrid web-OS attack scenarios and fundamental security vulnerabilities. Benchmarking current frontier CUAs identifies significant vulnerabilities: Claude 3.7 Sonnet | CUA demonstrates an Attack Success Rate (ASR) of 42.9%, while Operator, the most secure CUA evaluated, still exhibits an ASR of 7.6%. Notably, CUAs often attempt to execute adversarial tasks with an Attempt Rate as high as 92.5%, although failing to complete them due to capability limitations. Nevertheless, we observe concerning ASRs of up to 50% in realistic end-to-end settings, indicating that CUA threats can already result in tangible risks to users and computer systems. Overall, RedTeamCUA provides an essential framework for advancing realistic, controlled, and systematic analysis of CUA vulnerabilities, highlighting the urgent need for robust defenses to indirect prompt injection prior to real-world deployment.

Daniel Tan, Anders Woodruff, Niels Warncke, Arun Jose, Maxime Riché, David Demitri Africa, Mia Taylor

Language model finetuning often results in learning undesirable traits in combination with desired ones. To address this, we propose inoculation prompting: modifying finetuning data by prepending a short system-prompt instruction that deliberately elicits the undesirable trait. At test time, we evaluate without the instruction; inoculated models have much lower expression of the trait than models trained with unmodified training data. Inoculation is selective: in a toy setting where assistant responses are always in Spanish and ALL-CAPS, an appropriate inoculation (e.g., "You always speak in Spanish.") teaches the model to capitalize responses while still responding in English. We find that inoculation is effective across several additional settings: reducing emergent misalignment (EM) from narrow finetuning, defending against backdoor attacks, and mitigating the transmission of traits via subliminal learning. Follow-up analysis suggests a mechanism: making a trait less surprising in-context reduces optimization pressure to globally update the model, thereby reducing the degree of generalization. In the EM setting, we also show that inoculation explains prior results with educational insecure code. Beyond demonstrating a simple and effective technique for selective learning, our results contribute to a better conceptual understanding of how and why language models generalize.

Social Aspects · Privacy-preserving Statistics and Machine Learning

Linru Zhang, Xiangning Wang, Jun Jie Sim, Zhicong Huang, Jiahao Zhong, HUAXIONG WANG, Pu Duan, Kwok Yan Lam

Privacy concerns have been raised in Large Language Models (LLM) inference when models are deployed in Cloud Service Providers (CSP). Homomorphic encryption (HE) offers a promising solution by enabling secure inference directly over encrypted inputs. However, the high computational overhead of HE remains a major bottleneck. To address this challenge, we propose MOAI, an efficient HE-based, non-interactive framework for secure transformer inference. MOAI gains significant efficiency improvement from: (1) a novel evaluation flow that combines column and diagonal packing with consistent strategies across all layers, eliminating expensive format conversions. (2) rotation-free algorithms for Softmax and LayerNorm that significantly reduce the number of costly HE rotations, removing 2448 HE rotations in BERT-base inference. (3) Column packing removes rotations in plaintext–ciphertext matrix multiplications and interleaved batching further reduces the rotations in ciphertext–ciphertext matrix multiplications. MOAI uses at least 1.7x fewer HE rotations compared to the state-of-the-art works across all matrix multiplications of BERT-base. As a result, We achieve a 52.8\% reduction in evaluation time compared to the state-of-the-art HE-based non-interactive secure transformer inference, THOR (Moon et al., CCS'25). We then apply MOAI on the Powerformer's framework and achieve a 55.7\% reduction in evaluation time compared to Powerformer (Park et al., ACL'25), which approximates Softmax and LayerNorm with simpler functions in transformer and proposes HE-based non-interactive transformer inference. We report an amortized time of 2.36 minutes per input on a single GPU environment. We show the extendibility by applying MOAI in LLaMA-3-8B. Our implementation is publicly available as open source.

Xun Huang, Simeng Qin, Xiaoshuang Jia, Ranjie Duan, Huanqian Yan, Zhitao Zeng, Fei Yang, Yang Liu, Yang Liu

As Large Language Models (LLMs) are increasingly used, their security risks have drawn increasing attention. Existing research reveals that LLMs are highly susceptible to jailbreak attacks, with effectiveness varying across language contexts. This paper investigates the role of classical Chinese in jailbreak attacks. Owing to its conciseness and obscurity, classical Chinese can partially bypass existing safety constraints, exposing notable vulnerabilities in LLMs. Based on this observation, this paper proposes a framework, CC-BOS, for the automatic generation of classical Chinese adversarial prompts based on multi-dimensional fruit fly optimization, facilitating efficient and automated jailbreak attacks in black-box settings. Prompts are encoded into eight policy dimensions—covering role, behavior, mechanism, metaphor, expression, knowledge, trigger pattern and context; and iteratively refined via smell search, visual search, and cauchy mutation. This design enables efficient exploration of the search space, thereby enhancing the effectiveness of black-box jailbreak attacks. To enhance readability and evaluation accuracy, we further design a classical Chinese to English translation module. Extensive experiments demonstrate that effectiveness of the proposed CC-BOS, consistently outperforming state-of-the-art jailbreak attack methods.

Computer Vision · Vision Models & Multimodal

Fujun Han, Jingqi Ye, Chenglong Zhang, Peng Ye

Various vision intrusion detection models have achieved great success in many scenarios, e.g., autonomous driving, intelligent monitoring, and security, etc. However, their reliance on pre-defined classes limits their applicability in open-world intrusion detection scenarios. To remedy these, we introduce the Open-Vocabulary Intrusion Detection (OVID) project for the first time. Specifically, we first develop a novel dataset, Cityintrusion-OpenV for OVID, with more diverse intrusion categories and corresponding text prompts. Then, we design a multi-modal, multi-task, and end-to-end open-vocabulary intrusion detection framework named OVIDNet. It achieves open-world intrusion detection via aligning visual features with language embeddings. Further, two simple yet effective strategies are proposed to improve the generalization and performance of this specific task: (1) A Multi-Distributed Noise Mixing strategy is introduced to enhance the location information of unknown and unseen categories. (2) A Dynamic Memory-Gated module is designed to capture the contextual information under complex scenarios. Finally, comprehensive experiments and comparisons are conducted on multiple dominant datasets, e.g., COCO, Cityscape, Foggy-Cityscape, and Cityintrusion-OpenV. Besides, we also evaluate the universal applicability of our model in real scenarios. The results show that our method can outperform other classic and promising methods, and reach strong performance even under task-specific transfer and zero-shot settings, demonstrating its high practicality.

Applications · Genetics, Cell Biology, Health, etc

ZAIXI ZHANG, Zhenghong Zhou, Ruofan Jin, Le Cong, Mengdi Wang

DNA, encoding genetic instructions for almost all living organisms, fuels groundbreaking advances in genomics and synthetic biology. Recently, DNA Language Models have achieved success in designing synthetic functional DNA sequences, even whole genomes of novel bacteriophage, verified with wet lab experiments. Such remarkable generative power also brings severe biosafety concerns about whether DNA language models can design human viruses. With the goal of exposing vulnerabilities and informing the development of robust safeguarding techniques, we perform a systematic biosafety evaluation of DNA language models through the lens of jailbreak attacks. Specifically, we introduce JailbreakDNABench, a benchmark centered on high-priority human viruses, together with an end-to-end jailbreak framework, GeneBreaker. GeneBreaker integrates three key components: (1) an LLM agent equipped with customized bioinformatics tools to design high-homology yet non-pathogenic jailbreak prompts, (2) beam search guided by PathoLM and log-probability heuristics to steer sequence generation toward pathogen-like outputs, and (3) a BLAST- and function-annotation–based evaluation pipeline to identify successful jailbreaks. On JailbreakDNABench, GeneBreaker successfully jailbreaks the latest Evo series models across 6 viral categories consistently (up to 60\% Attack Success Rate for Evo2-40B). Further case studies on SARS-CoV-2 spike protein and HIV-1 envelope protein demonstrate the sequence and structural fidelity of jailbreak output, while evolutionary modeling of SARS-CoV-2 underscores biosecurity risks. Our findings also reveal that scaling DNA language models amplifies dual-use risks, motivating enhanced safety alignment and tracing mechanisms.

Yuyang Ding, Chi Zhang, Juntao Li, Haibin Lin, Xin Liu, Min Zhang

Reinforcement learning with verifiable rewards (RLVR) has emerged as a promising paradigm for enhancing the reasoning capabilities of large language models (LLMs). In this context, models explore reasoning trajectories and exploit rollouts with correct answers as positive signals for policy optimization. However, these rollouts might involve flawed patterns such as answer-guessing and jump-in-reasoning. Such flawed-positive rollouts are rewarded identically to fully correct ones, causing policy models to internalize these unreliable reasoning patterns. In this work, we first conduct a systematic study of flawed-positive rollouts in RL and find that they enable rapid capability gains during the early optimization stage, while constraining reasoning capability later by reinforcing unreliable patterns. Building on these insights, we propose **F**lawed-**A**ware **P**olicy **O**ptimization (**FAPO**), which presents a parameter-free reward penalty for flawed-positive rollouts, enabling the policy to leverage them as useful shortcuts in the warm-up stage, securing stable early gains, while gradually shifting optimization toward reliable reasoning in the later refinement stage. To accurately and comprehensively detect flawed-positive rollouts, we introduce a generative reward model (GenRM) with a process-level reward that precisely localizes reasoning errors. Experiments show that FAPO is effective in broad domains, improving outcome correctness, process reliability, and training stability without increasing the token budget.

Reinforcement Learning · Multi-agent

Wei Du, Benyu Wu, YUQING SUN, Wei Guo, yuntao du, Zhongmin Yan, Guoxian Yu, Lizhen Cui

Effective communication is crucial for multi-agent cooperation in partially observable environments. However, a fundamental trilemma exists among task performance, communication efficiency, and human interpretability. To resolve this, we propose a multi-agent communication framework via $\textbf{G}$rounding $\textbf{L}$anguage and $\textbf{C}$ontrastive learning (GLC) to learns efficient and interpretable communication protocols. Specifically, GLC employs an autoencoder to learn discretized and compressed communication symbols, ensuring high communication efficiency. These symbols are then semantically aligned with human concepts using data generated by a Large Language Model (LLM), making them human-interpretable. Furthermore, a contrastive learning objective is introduced to ensure consistency and mutual intelligibility among all agents, thereby securing high task utility. GLC dynamically balances these objectives by the Information Bottleneck principle. Extensive experiments show that GLC outperforms state-of-the-art methods across multiple benchmarks, delivering superior task performance, higher communication efficiency, and enhanced human interpretability.