EgoPoint-Bench - Simulation Adversarial Resilience: leaderboard
EgoPoint-Bench (2026) asks multimodal models about the object a first-person hand points at, on held-out simulation scenes and real smart-glasses recordings, zero-shot; true/false and single-choice answers by exact match and open answers scored by a GPT-4o judge. This board covers the Adversarial Resilience questions (counterfactual, fallacious or void references) among the 1,929 simulation test questions. Accuracy in percent; higher is better.
Metric: Accuracy (%) on the Adversarial Resilience questions among the 1,929 held-out simulation test questions (HM3D, HSSD and AI2-THOR scenes); zero-shot (direct inference); true/false and single-choice questions by exact match, open-ended answers scored 0 to 1 by a GPT-4o judge; greedy decoding for open models, temperature 0 for API models; higher is better. Source: arxiv.org. Saturation forecast: Around 2029. 12 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 3 VL 32B Instruct | 60.23 | #606 |
| 2 | Gemini 3 Flash | 58.39 | #175 |
| 3 | Gemini 3 Pro | 57.16 | #145 |
| 4 | Qwen 3 VL 8B Instruct | 52.11 | #1029 |
| 5 | GPT-5.2 Instant | 50.88 | #304 |
| 6 | LLaVA-NeXT-7B | 46.59 | #5032 |
| 7 | LLaVA-v1.5-7B | 41.92 | #7630 |
| 8 | GLM-4.6V-Flash | 36.17 | #1040 |
| 9 | GPT-5 Mini | 36.09 | #364 |
| 10 | InternVL3.5-8B-Instruct | 35.63 | #1891 |
| 11 | InternVL3.5-14B-Instruct | 33.56 | #1736 |
| 12 | InternVL3.5-2B-Instruct | 26.05 | #6517 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=egopoint-bench-simulation-adversarial-resilience · How It Works · Data refreshed daily, snapshot 2026-11-21.