TelemetrySuffBench - Origin-Step Localization: leaderboard

Metric: Origin-step Top-1 accuracy (%; the model names the exact event that introduced the fault; over TelemetrySuffBench's 216 delayed-binding fault traces; full-telemetry view; closed candidate sets, invalid outputs scored zero). Source: arxiv.org. Saturation forecast: Around December 2026. 5 models tracked.

Top models

#ModelScore
1GPT-5.6 Sol97.2
2Claude Opus 597.2
3GPT-5.590.3
4DeepSeek V4 Pro84.7
5Qwen 3.7 Max33.8

Interactive version: theaggregate.ai/benchmark?slug=telemetrysuffbench-origin-step-localization · How It Works · Data refreshed daily, snapshot 2026-09-26.