Boolean Satisfiability (SAT) on SAT Easy (test)
98AccuracyRecursive Model
Evaluation Results
| Method | Links | |
|---|---|---|
| Recursive ModelBackbone=Qwen2.5-3B-Instruct, Framework=Recursive2026.03 | 98 | |
| Qwen3-235B2026.03 | 88 | |
| DeepSeek-Distill-14B2026.03 | 84.3 | |
| GPT-4o2026.03 | 69.9 | |
| LLaMA3.3-70B2026.03 | 65.1 | |
| Random Baseline2026.03 | 50 |