ReaLM - AdvEDM: leaderboard
Metric: Attack success rate (%; share of adversarial cases where the model gives an incorrect, task-relevant answer) under AdvEDM-A (semantic editing that steers attention toward the target region) on the 832 ReaLM physical-world cases (driving, manipulation, grasping, physics, egocentric, diagnostics and scene QA; image, question and ground-truth answer), black-box one-shot evaluation at temperature 0, a Qwen3-8B judge extracting the answer; lower is better. Source: arxiv.org. Saturation forecast: Around September 2027. 13 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3 Flash | 36.5 |
| 2 | Qwen 3.5 27B | 38.8 |
| 3 | Qwen 3.5 122B A10B | 39.3 |
| 4 | Qwen 3.6 Flash | 40.9 |
| 5 | Kimi K2.5 | 41.6 |
| 6 | Qwen 3 VL 32B | 43.1 |
| 7 | GPT-4.1 Mini | 43.5 |
| 8 | Qwen 3.5 9B | 43.9 |
| 9 | Seed 2.0 Mini | 44.1 |
Interactive version: theaggregate.ai/benchmark?slug=realm-advedm · How It Works · Data refreshed daily, snapshot 2026-09-29.