FoundationalASSIST - Cognitive Student Modeling: leaderboard
Metric: Exact-answer accuracy (%; the option or value the student will enter, same predictions; numeric answers within 1%). Source: arxiv.org. Saturation forecast: Around February 2028. 4 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Llama 3.3 70B Instruct | 44.3 |
| 2 | GPT-OSS-120B | 41.2 |
| 3 | Qwen 3 Next 80B A3B (Thinking) | 38.5 |
| 4 | Qwen 3 Next 80B A3B Instruct | 35.5 |
Interactive version: theaggregate.ai/benchmark?slug=foundationalassist-cognitive-student-modeling · How It Works · Data refreshed daily, snapshot 2026-09-26.