Medical Reasoning on MedQA (EM)
61.1EMGraphWalker
Evaluation Results
| Method | Links | |
|---|---|---|
| GraphWalkerBackbone LLM=Qwen3-14B2026.04 | 61.1 | |
| Semantic-embBackbone LLM=Qwen3-14B2026.04 | 59.58 | |
| CONEBackbone LLM=Qwen3-14B2026.04 | 59.17 | |
| Zero-shotBackbone LLM=Qwen3-14B2026.04 | 56.56 | |
| SPELLBackbone LLM=Qwen3-14B2026.04 | 54.1 | |
| LMS3Backbone LLM=Qwen3-14B2026.04 | 54.09 | |
| Delta-KNNBackbone LLM=Qwen3-14B2026.04 | 53.47 | |
| InfluenceBackbone LLM=Qwen3-14B2026.04 | 53.02 | |
| RandomBackbone LLM=Qwen3-14B2026.04 | 52.96 | |
| GradSelBackbone LLM=Qwen3-14B2026.04 | 52.93 | |
| IDSBackbone LLM=Qwen3-14B2026.04 | 52.59 | |
| Time-seriesBackbone LLM=Qwen3-14B2026.04 | 52.11 |