GeoThinker-F Study - Vanilla - Room size: leaderboard
S3.T1; new GeoThinker-F trained checkpoints only; Vanilla training regime retained as distinct identity; quoted baseline literature scores omitted; candidate coverage is not fitted eligibility. Metric: Accuracy (%). Higher is better. Source: https://arxiv.org/html/2610.20592v1
Metric: Accuracy (%). Source: arxiv.org. 1 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GeoThinker-F-Qwen2.5VL-7B (Vanilla) | 58.1 |
Interactive version: theaggregate.ai/benchmark?slug=geothinker-f-study-vanilla-room-size · How It Works · Data refreshed daily, snapshot 2026-10-16.