PatEx-Gym: First Agent Benchmark for Patent Examination with 1M+ Patent Corpus
IPIntelligence · hf · 2026-10-27
Researchers from the Shenzhen Institute of Advanced Technology (CAS), Dalian University of Technology, and partners present PatEx-Gym, the first framework for developing and evaluating agents for patent examination, covering the full search–reasoning–assessment pipeline for novelty assessment.
- PatEx-Bench: 1K realistic examination tasks, two large-scale retrieval corpora with over 1M patents and 52K scientific papers, and a harness runtime supporting iterative search, document visits, reasoning, and Office Action generation.
- PASS: an annotation-free evaluation approach for novelty conclusions.
- Includes synthetic trajectory data and a compact agent for iterative prior-art search.
The authors argue existing benchmarks focus on static, limited candidate sets, leaving the full examination process underexplored.
More from Research
- AI-designed drug works in humans: aging looks 'like a list of fixable problems' — rand_longevity · 2026-10-27
- Pessimistic complexity theory hypothesis: faster algorithms must get uglier — Elijah_Meeks · 2026-10-27
- UMAP arcing explained: one giant coordinate dominates neural network activation maps — leland_mcinnes · 2026-10-27
- NeurIPS benchmark: tree models still beat deep learning on 45 tabular datasets — tak3sh8 · 2026-10-27
- Kevin Bryan: maximizing raw accuracy is rarely the right way to train AI assistants — Afinetheorem · 2026-10-27
- Szegedy: a flaw in common training losses went unnoticed for over a decade — ChrSzegedy · 2026-10-27