AgentAbstain: Do LLM Agents Know When Not to Act?
cs.AI
Submitted: 2026-07-11
Updated: 2026-11-18
Terminology
Sources
- AgentHarm: A Benchmark for Measuring Harmfulness of LLM Agents
- Defensive Refusal Bias: How Safety Alignment Fails Cyber Defenders
- The BrowserGym Ecosystem for Web Agent Research
- Just Enough Shifts: Mitigating Over-Refusal in Aligned Language Models with Targeted Representation Fine-Tuning
- DeepSeek-V3.2: Pushing the Frontier of Open Large Language Models
- HiL-Bench (Human-in-Loop Benchmark): Do Agents Know When to Ask for Help?
- ClarQ-LLM: A Benchmark for Models Clarifying and Requesting Information in Task-Oriented Dialog
- GLM-5: from Vibe Coding to Agentic Engineering
- Coding Agents Don't Know When to Act
- Are Your Agents Upward Deceivers?
- OpaqueToolsBench: Learning Nuances of Tool Behavior Through Interaction
- DeceptionBench: A Comprehensive Benchmark for AI Deception Behaviors in Real-world Scenarios
- Calibrating Verbal Uncertainty as a Linear Feature to Reduce Hallucinations
- Large Language Models Must Be Taught to Know What They Don't Know
- MCPVerse: An Expansive, Real-World Benchmark for Agentic Tool Use
- ST-WebAgentBench: A Benchmark for Evaluating Safety and Trustworthiness in Web Agents
- AgentBench: Evaluating LLMs as Agents
- ToolSandbox: A Stateful, Conversational, Interactive Evaluation Benchmark for LLM Tool Use Capabilities
- Agentic Abstention: Do Agents Know When to Stop Instead of Act?
- ClarifyMT-Bench: Benchmarking and Improving Multi-Turn Clarification for Conversational Large Language Models
Related papers
- Multi-Objective Structured Pruning of LLMs for Latency and Model Size Optimization
- MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
- Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
- Embeddings for Preferences, Not Semantics
- Can We Trust Item Response Theory for AI Evaluation?
- Diagnosing Task Insensitivity in Language Agents