PAL-Bench (PAL-TRACE) - Owner Fact Recovery: leaderboard
Metric: Owner-fact recovery (0-1; share of a user's hidden owner-fact atoms, 840 in all, whose content a predicted fact matches; PAL-TRACE framework with the listed LLM as reconstruction backbone; 50 synthetic users, macro-averaged over users; Qwen3.6-35B-A3B semantic judge). Source: arxiv.org. Saturation forecast: Around June 2027. 9 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 0.61 |
| 2 | GPT-5.4 | 0.61 |
| 3 | Claude Sonnet 4.5 | 0.6 |
| 4 | Gemini 3.1 Flash Lite | 0.6 |
| 5 | DeepSeek V4 Flash | 0.58 |
| 6 | GPT-5.4 Mini | 0.56 |
| 7 | Qwen 3.6 35B A3B | 0.54 |
| 8 | Gemma 4 26B A4B (IT) | 0.5 |
Interactive version: theaggregate.ai/benchmark?slug=pal-bench-pal-trace-owner-fact-recovery · How It Works · Data refreshed daily, snapshot 2026-09-26.