TAR-Bench - Causal Linkage: leaderboard

Metric: Scaled BERTScore F1 (%; an explanation of what caused the anomaly, compared with the reference; zero-shot in non-reasoning (direct-answer) mode through VLMEvalKit, mean of 3 runs, on TAR-Bench: 960 human-curated annotations for 80 held-out traffic-anomaly clips from 17 public YouTube videos). Source: arxiv.org. Saturation forecast: Around 2032. 11 models tracked.

Top models

#ModelScore
1Gemma 4 31B (IT)27.9
2Qwen 3.5 27B (Non-reasoning)26.6
3Qwen 3 VL 8B Instruct24.3
4Gemini 3.1 Pro (Preview)23.5
5Qwen 3 VL 32B Instruct17.4
6Qwen 3.5 9B (Non-reasoning)17

Interactive version: theaggregate.ai/benchmark?slug=tar-bench-causal-linkage · How It Works · Data refreshed daily, snapshot 2026-09-29.