Another 201papers do not have category tags yet and are not included in filtered results.
Proposes RATED, which reranks agent tool retrieval using execution feedback
Proposes SWE-DeceptionBench to evaluate LLM deception in the software development lifecycle
Auditing training labels for label-only backdoors, comparing spaces and statistics, and judging by retraining
The authors ask: can a defender without trusted data decide which training labels to keep, and how much clean data this destroys, and what the final model accuracy is.
Introduces DistressBench, evaluating models' multidimensional support in crisis conversations
DistressBench uses SSH dialogues written and revised by clinical personnel, and evaluates how much necessary care general chat models deliver according to weighted atomic criteria, rather than whether they refuse. When users disclose suicidal ideation or self-harm to a chat model, the authors consider the expected behavior to be recognizing the crisis, calibrating empathy, de-escalating, and referring to human help. Refusal-based evaluations cannot detect failing to act after recognizing the crisis.
Measuring the impact of random weight bit flips on LLM agent safety and alignment behavior
After random two-bit weight faults enter an LLM agent, model-level damage becomes recoverable, increased cost, subtle semantic errors, and changes in safety or alignment behavior.
Proposes the Cancerous backdoor, which progressively corrupts reasoning using invisible stage tokens.
Cancerous Backdoor Attack changes LLM backdoors from one-time activation to multi-stage reasoning corruption that accumulates progressively with triggers.
Proposes Task Bank self-replay to mitigate held-out forgetting in continual instruction tuning
The authors study how language models, when learning new tasks, can use their own knowledge for broader self-preservation: given an incomplete list and forgetting that extends beyond the list, can the model generate replay and help choose targets.
Proposes the ISE framework to evaluate the epistemic humility of LLM agents under knowledge conflicts
What the authors want to test is not whether the agent's final answer is correct, but whether, when a knowledge conflict arises, it recognizes the gap in its trajectory, performs bounded tool follow-up, and writes the unresolved uncertainty to the user.
Diagnose whether compliance system rulings change with perturbations to regulatory rules
Use behavioral counterfactuals and activation probes to audit whether compliance rulings rely on the given rules. The authors aim to separate two kinds of seemingly identical correct rulings: those that use the given rules, and those obtained from case surface features or training patterns.
Analyzing why optimizing teaching adaptation metrics collapses tutoring choices and diverges from expert judgment
After turning teaching adaptation metrics into a fine-tuning objective, the metrics rise while scores from trained educators decline.
Proposes LGWM, using an action-conditioned world model to verify GUI agent screen transitions.
评测编码 Agent 中相似技能共同安装对已装技能核心功能的替代
作者对共同安装、做同一件事的良性 skill 做实证研究，用核心功能是否还在，而不是任务过不过，来看冲突。。
测量 LLM 版本更替对科学摘要 AI 改写检测器的影响
作者量化的是：科学摘要的 AI 改写筛查，在 LLM 版本不断更换、而人类摘要误标必须少见时，会怎样变。问题来自期刊和会议已经按分数阈值筛查投稿，但基准多停在评估时可得的版本上。
提出 SSCBench，评测工具调用 Agent 故障注入测试的证据有效性
SSCBench 研究工具使用智能体的故障注入结果，在什么证据条件下可以被解释。
提出 RL-ARC，用推理置信度在强化学习中校准推理模型
RL-ARC 是校准感知的强化学习框架，用推理过程的口头置信度调节答案置信度。
用 BeliefScope 分离固定命题上证据与压力对模型信念的影响
BeliefScope 是一个黑盒行为诊断框架，用来把固定命题上的回答变化拆成证据驱动的修正，和不加新事实的方向性压力。信念修正评测能看到信息基础变化后的更新，谄媚评测能看到用户影响下的顺从或漂移。作者认为，当两种输入都偏向同一结论时，表面偏移本身不能区分来源。
构建 EpiReal-Bench，评测图像生成器将虚假主张渲染为可信证据
EpiReal-Bench 用来量商用图像生成器会不会把虚假现实主张画成可信视觉证据。
用时钟 harness 和预算塑形 GRPO 改善 Agent 的按时返回
作者研究小规模 LLM 智能体在显式 wall-clock 预算下能否按时返回，并把更多时间换成更好的解。
提出 PatchBench-Local 协议，评测激活补丁修复越狱的局部附带损伤
Abstract论文提出 PatchBench 与局部评估协议，揭示了激活补丁在全局能力无损下仍存局部良性退化。
测量工具调用 Agent 对显式报错与静默损坏的检测和恢复
作者在 BFCL 多轮可执行环境上注入一次工具层故障，把注意到、改计划、恢复和重复分开记分，而不是只看任务有没有做完。