Logical Fallacy Tutoring on Elec2Deb20 adversarial student
11.9Divergence RateBASE W/ PROBLEMS
Evaluation Results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| BASE W/ PROBLEMSBackbone=GPT-4o, Evaluator=QwQ-32B, Student Type=adversarial2026.05 | 11.9 | 37.3 | 27 | 15.4 | 49.5 | 40.3 | 30.2 | |
| LFTutorBackbone=GPT-4o, Evaluator=QwQ-32B, Student Type=adversarial2026.05 | 38 | 84 | 82 | 32.5 | 86.2 | 41.2 | 60.7 |