DiagChain - Noisy Evidence: leaderboard

Metric: Correctly reconstructed reference steps (%; share of the 283 coverable reference steps of the 23 noisy-evidence cases (sampled benign or distracting activity added); attack-chain reconstruction from heterogeneous telemetry with the ECRAG evidence-retrieval scaffold; a step counts when its evidence is observed, used, fully attributed and correctly ordered, with no earlier failure; temperature 0). Source: arxiv.org. Saturation forecast: Around October 2027. 6 models tracked.

Top models

#ModelScore
1GPT-5.541.7
2GLM-5.2 (High)26.5
3GLM-5.2 (Non-reasoning)24.03
4DeepSeek V4 Pro (Non-reasoning)18.73

Interactive version: theaggregate.ai/benchmark?slug=diagchain-noisy-evidence · How It Works · Data refreshed daily, snapshot 2026-09-29.