Another 240papers do not have category tags yet and are not included in filtered results.
Proposes RATED, which reranks agent tool retrieval using execution feedback
Propose X-AgentRed, which uses a tool dependency graph to red-team cross-service unauthorized harm in personal agents
Proposes SAFE-QE, which filters unsupported hallucinated words in LLM query expansion
AbstractSAFE-QE suppresses expansion hallucinations with reproducibility and coverage selection; the authors say ranking quality shows no visible cost.
Proposes SWE-DeceptionBench to evaluate LLM deception in the software development lifecycle
Proposes DREAM, which selectively weakens agent-poisoned memories while preserving benign memories.
Proposing DRYLAB-Bench to evaluate agentic LLMs' high-risk mutation design capability
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.
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 PASAC and a five-layer boundary guarantee stack, turning agent evaluation overstepping into verifiable whole-system guarantees.
Proposes the FAITH safe filtering reinforcement learning framework, training high-dimensional systems to optimize task reward under constraints
FAITH is a model-free safe filtering reinforcement learning framework for high-dimensional systems: safety correction and task reward reside in separate modules.
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.
Proposing an option channel attack that renames option labels so that agent guardrails allow prohibited operations
AbstractThe authors' evaluation indicates that typed models should not serve as final safety arbiters; they are susceptible to off-policy text and option name manipulation, and confidence filtering cannot resist attacks.
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.
Using fear activation steering to test the hazard judgment criteria of vision-language models
Proposes LGWM, using an action-conditioned world model to verify GUI agent screen transitions.
Measures citation fragility in AI search, demonstrating that ordinary public posts can enter citations and answers.
AbstractThe authors claim the paper quantifies vulnerabilities in AI search caused by citation preferences and low posting thresholds; tests show public web posts can penetrate AI citations, and they call for auditing the citation supply chain.
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。。