ATRBench - Oracle Rules: leaderboard
Metric: ATRBench test-session accuracy TSAcc (%, 0-100): the share of user-offline rule-bound test sessions whose first relevant tool decision matches the gold action of the bound standing rule (a deterministic check on the trajectory), macro-averaged over 20 synthetic personas (284 rule and test-session pairs in six personal-assistant tool domains); learning sessions with a simulated user run first and their transcript is frozen as context for the test sessions; single trial, 20-turn cap, provider reasoning modes enabled; oracle variant (oracle_target): learning is skipped and the canonical answer of the test-bound rule is supplied as a prior user statement, the application ceiling; higher is better. Source: arxiv.org. Saturation forecast: Estimated already saturated. 8 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 96.7 |
| 2 | Qwen 3.6 Plus (Thinking) | 93 |
| 3 | DeepSeek V4 Flash (High) | 93 |
| 4 | Claude Opus 4.7 (Thinking) | 92.5 |
| 5 | DeepSeek V4 Pro (High) | 89.5 |
| 6 | GPT-5.4 (High) | 88.6 |
| 7 | MiniMax-M2.7 | 82.5 |
Interactive version: theaggregate.ai/benchmark?slug=atrbench-oracle-rules · How It Works · Data refreshed daily, snapshot 2026-10-07.