FIRE-Bench - Cross-Domain: leaderboard

Metric: Claim-level F1 (%; atomic-claim entailment against the source findings with a GPT-5.2 judge; mean over the 10 cross-domain tasks (five CV/VLM papers, five neural-network-analysis papers) and three runs). Source: arxiv.org. Saturation forecast: Rough model projection: around 2026. 3 models tracked.

Top models

#ModelScore
1GPT-5 (Codex CLI, medium)30.2
2o4-mini (OpenHands)23.4
3GPT-5 (OpenHands)13.3

Interactive version: theaggregate.ai/benchmark?slug=fire-bench-cross-domain · How It Works · Data refreshed daily, snapshot 2026-09-26.