ImmersedPrivacy - Inferred Privacy Planning (Native Audio): leaderboard
History-conditioned privacy adherence in ImmersedPrivacy (Wang et al., 2026), Tier 3: the model watches a simulated video, with the observed dialogue as audio, in which a person hides an item and asks that it stay private, then gets a conflicting command and selects which of three candidate actions to perform. Exact match in percent (both task actions chosen, the privacy-violating one avoided); higher is better.
Metric: Exact match (%; Tier 3: share of cases where the model selects exactly the two task actions that leave alone the item an observed person wanted kept private; social-state audio or dialogue given as raw audio). Source: arxiv.org. Saturation forecast: Around October 2027. 6 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) (High) | 51 | #91 (Gemini 3.1 Pro (Preview)) |
| 2 | Gemini 3.1 Pro (Preview) (Low) | 46 | #91 (Gemini 3.1 Pro (Preview)) |
| 3 | Gemini 3 Flash (Preview) (High) | 15 | #141 (Gemini 3 Flash (Preview)) |
| 4 | Gemini 3 Flash (Preview) (Low) | 9 | #141 (Gemini 3 Flash (Preview)) |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=immersedprivacy-inferred-privacy-planning-native-audio · How It Works · Data refreshed daily, snapshot 2026-11-12.