CASE - Frozen DFEW WAR: leaderboard
CASE emotion paper v1, frozen-backbone baseline cohort only. CA-MER audio/visual-conflict and consistent subsets, plus DFEW weighted/unweighted recall. No skill editor or CASE intervention is attributed to the baseline model. Evolution/validation/test sets disjoint. Qwen3-VL checkpoint variant and Emotion-Qwen release not bound in the paper and retained source-qualified; Gemini-3.5-Flash effort unspecified. DFEW WAR. Metric: Recall (%).
Metric: Recall (%). Source: arxiv.org. Saturation forecast: Around March 2027. 3 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3.5 Flash | 60.92 | #64 |
No result here: #3 Claude Opus 5.5, #5 GPT-6 Astra, #6 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=case-frozen-dfew-war · How It Works · Data refreshed daily, snapshot 2026-10-27.