Equation-Suffix Prediction (Qwen3-8B Scorer): leaderboard
Metric: Likelihood lift (mean clipLL2 per target token over the same-budget recent-context control, range -2 to 2; 1,363 equation-suffix cuts from 138 recent arXiv physics and mathematics papers; Qwen3-8B scorer). Source: arxiv.org. Saturation forecast: Around November 2027. 9 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.5 (High) | 0.2 |
| 2 | GPT-5.5 (Medium) | 0.2 |
| 3 | GPT-5.5 (Low) | 0.19 |
| 4 | GPT-5.5 (Non-reasoning) | 0.17 |
| 5 | GPT-5.4 Nano (High) | 0.11 |
| 6 | GPT-5.4 Nano (Medium) | 0.1 |
Interactive version: theaggregate.ai/benchmark?slug=equation-suffix-prediction-qwen3-8b-scorer · How It Works · Data refreshed daily, snapshot 2026-09-26.