Research map / Multi-agent systems and games / Human-AI and goal-conditioned
AutoLibra: Agent Metric Induction from Open-Ended Human Feedback
Hao Zhu, Phil Cuvin, Xinkai Yu, Charlotte Ka Yee Yan, Jason Zhang, Diyi Yang
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| keep concrete good and bad behavior examples in induced metrics (ablation) | AutoLibra · CoGym | coverage: up to 30% coverage decrease when dropped | Improved |
| evaluate coverage on held-out trajectories instead of induction trajectories | AutoLibra · four agentic datasets | coverage: <5% lower | No clear effect |
| use K-means on aspect embeddings instead of LLM clustering (ablation) | behavior clustering · agentic datasets | cluster quality | Hurt |
| use genetic algorithm or iterative clustering for metric optimization (ablation) | AutoLibra · agentic datasets | coverage/redundancy | No clear effect |
| induce new metrics from scratch each iteration instead of incrementally (ablation) | AutoLibra · agentic datasets | coverage | No clear effect |
Closest papers by meaning
- Agent-as-a-Judge: Evaluate Agents with Agents (ICML 2025 · similarity 0.89)
- Talk, Evaluate, Diagnose: User-aware Agent Evaluation with Automated Error Analysis (ICLR 2026 · similarity 0.89)
- UniClawBench: A Universal Benchmark for Proactive Agents on Real-World Tasks (NeurIPS 2026 · similarity 0.88)
- Claw-Eval: Towards Trustworthy Evaluation of Autonomous Agents (NeurIPS 2026 · similarity 0.88)
- Proactive Agent: Shifting LLM Agents from Reactive Responses to Active Assistance (ICLR 2025 · similarity 0.87)
- AOrchestra: Automating Sub-Agent Creation for Agentic Orchestration (ICML 2026 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
