FireWorldBench - Counterfactual Intervention Reasoning (Rendered Images): leaderboard
Metric: Completion accuracy (%; P5, counterfactual intervention reasoning items of the controlled-simulation subset; image-observation track (physics-rendered field visualizations and 3D scene renderings); each cell is the mean of the matched choice-format score (Jaccard overlap of selected and gold option sets) and open-report score (exact agreement of required structured slots); deterministic evaluator, no LLM judge; failed or unparseable responses score 0). Source: arxiv.org. Saturation forecast: Around 2030. 10 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 2.5 Flash | 43.46 |
| 2 | InternVL3-78B | 41.22 |
| 3 | Qwen 3.5 Plus (2026-02-15) | 37.83 |
| 4 | Qwen 3 VL 8B Instruct | 36.42 |
| 5 | Qwen 3 VL 32B Instruct | 35.82 |
| 6 | GPT-5.6 Luna | 34.61 |
| 7 | Claude Haiku 4.5 | 32.47 |
| 8 | InternVL3-38B | 31.94 |
| 9 | InternVL3-8B | 26.27 |
| 10 | Qwen 3 VL 235B A22B Instruct | 20.62 |
Interactive version: theaggregate.ai/benchmark?slug=fireworldbench-counterfactual-intervention-reasoning-rendered-images · How It Works · Data refreshed daily, snapshot 2026-09-26.