Trapped in Your Story - Full Narrative: leaderboard
Same 500 questions, counting only the strongest condition, where the user's position comes with anecdote, causal story and worldview framing. Isolates the effect of a full personal narrative from the any-level headline.
Metric: Stance reversal under the full-narrative condition L5 (%); lower is better. Source: arxiv.org. 7 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Fable 5 | 8.54 | #27 |
| 2 | Gemini 3.8 Flash | 11 | #36 |
| 3 | GPT-5.5 | 42 | #49 |
| 4 | GPT-5.1 | 54.11 | #229 |
| 5 | DeepSeek V4 Pro (Non-reasoning) | 92.38 | #188 (DeepSeek V4 Pro) |
Interactive version: theaggregate.ai/benchmark?slug=trapped-in-your-story-full-narrative · How It Works · Data refreshed daily, snapshot 2026-11-13.