Research map / Large language models / Language, safety and interpretability
Extracting Events Like Code: A Multi-Agent Programming Framework for Zero-Shot Event Extraction
Quanjiang Guo, Sijie Wang, Jinchuan Zhang, Ben Zhang, Zhao Kang, Ling Tian, Ke Yan
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 |
|---|---|---|---|
| add Retrieval Agent exemplar generation (ablation) | AEC · ACE 2005 trigger identification | F1: 49.8 -> 57.0 (Llama3-70B) | Improved |
| add planning rationales (ablation) | AEC · ACE 2005 trigger identification | F1: 52.6 -> 57.0 (Llama3-70B) | Improved |
| add verification loop (ablation) | AEC · ACE 2005 argument classification | F1: 28.5 -> 34.7 (Llama3-70B) | Improved |
| add structural check (ablation) | AEC · FewEvent trigger identification | F1: 39.8 -> 42.1 (Llama3-70B) | Improved |
| replace baselines with AEC multi-agent code framework | Llama3-70B · ACE 2005 trigger identification | F1: 51.5 -> 57.0 (best baseline CEDAR) | Improved |
| replace DecomposeEE with AEC | GPT3.5-turbo · ACE 2005 argument classification | F1: DecomposeEE 27.4 vs AEC 29.8 on GENIA AC, but CASIE AC 27.6 vs 26.7 | Mixed |
| increase hypotheses k from 1 to 3 (t=3) (ablation) | AEC with GPT4o · ACE 2005 event identification | F1: 56.8 -> 61.8 | Improved |
| increase k and t from 3 to 5 (ablation) | AEC with GPT4o · FewEvent / ACE | F1: 44.6 -> 45.1 (FewEvent TI), 58.3 -> 58.3 (ACE TI) | No clear effect |
| use more than three verification test cases (ablation) | AEC with GPT4o · FewEvent, ACE | score | No clear effect |
Closest papers by meaning
- ShortcutsBench: A Large-Scale Real-world Benchmark for API-based Agents (ICLR 2025 · similarity 0.85)
- Toward Interactive Understanding of Code APIs (NeurIPS 2026 · similarity 0.85)
- Kimi-Dev: Agentless Training as Skill Prior for SWE-agents (ICLR 2026 · similarity 0.85)
- PROACT-Agent: Progressive Runtime Oversight and Active Circuit-breaking for Real-Time Safety (NeurIPS 2026 · similarity 0.85)
- Executable Code Actions Elicit Better LLM Agents (ICML 2024 · similarity 0.84)
- Nemotron-CORTEXA: Enhancing LLM Agents for Software Engineering Tasks via Improved Localization and Solution Diversity (ICML 2025 · similarity 0.84)
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.
