LatentGUIWorld - Dynamics-Dependent (I2) Average: leaderboard
The four LatentGUIWorld environments (arXiv 2610.16720) whose hidden movement dynamics can only be identified from the effects of the agent's own actions: egocentric Drag, egocentric Ten-Choice, and inner- and outer-ring Rotation, 150 episodes each, at most 12 model responses per episode. Score: episode success rate (%), macro-average over the four. Higher is better. Source: https://arxiv.org/abs/2610.16720
Metric: Episode success rate (%; macro-average over the four environments whose hidden movement dynamics must be identified from the agent's own actions: egocentric Drag, egocentric Ten-Choice, inner and outer Rotation). Source: arxiv.org. Saturation forecast: Around February 2028. 11 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-6 Astra | 45 | #6 |
| 2 | Claude Opus 5 | 12.7 | #12 |
| 3 | Kimi K3 | 10.2 | #30 |
| 4 | GPT-5.6 Sol | 7.2 | #20 |
| 5 | uitars-1.5-7B | 1.7 | #863 |
| 6 | DeepSeek V4.1 Flash | 0.7 | #35 |
No result here: #3 Claude Opus 5.5, #7 GPT-6.1 Sol, #8 Gemini 4 Argon.
Interactive version: theaggregate.ai/benchmark?slug=latentguiworld-dynamics-dependent-i2-average · How It Works · Data refreshed daily, snapshot 2026-10-19.