RPCBench - Domain - Amazon Sports - Evidence Distortion: leaderboard
RPCBench tests proactive premise critique in recommendation requests. This series measures evidence distortion on the Amazon Sports subset (1,270 questions), using the same eleven-model, three-judge experiment as the headline paper. Published F1R fractions from the authors' per-model CSV are scaled once to Evidence Distortion Rate (%). Lower is better; domains and nested premise-error groups are related slices of the same experiment.
Metric: Evidence Distortion Rate (%). Source: github.com. Saturation forecast: Around February 2027. 11 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.5 | 12.68 | #34 |
| 2 | Gemini 3.1 Pro (Preview) | 14.8 | #66 |
| 3 | Qwen 3.5 Plus | 15.98 | #136 |
| 4 | Qwen 3.5 397B A17B | 16.93 | #155 |
| 5 | Claude Sonnet 4.6 | 17.72 | #97 |
| 6 | Qwen 3.5 122B A10B | 17.8 | #186 |
| 7 | Qwen 3.5 35B A3B | 19.45 | #270 |
| 8 | DeepSeek V4 Flash | 23.62 | #156 |
| 9 | Llama 3.1 8B Instruct | 24.57 | #1021 |
| 10 | DeepSeek V4 Pro | 25.51 | #113 |
| 11 | Llama 3.1 70B Instruct | 27.87 | #570 |
No result here: #3 Claude Opus 5.5, #5 GPT-6 Astra, #6 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=rpcbench-domain-amazon-sports-evidence-distortion · How It Works · Data refreshed daily, snapshot 2026-11-07.