Music Analysis on SSMR Music (test)
96Scale ScoreSkill-based LLM agent
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| Skill-based LLM agentType=Adaptive, Backbone=Qwen3-4B, Iterations=162026.06 | 96 | 88 | 74.4 | 61.6 | 80 | |
| LLMZEROType=Adaptive, Backbone=Qwen3-4B, Iterations=162026.06 | 94.4 | 81.6 | 77.6 | 75.2 | 82.2 | |
| Random searchType=Static, Backbone=Qwen3-4B, Iterations=162026.06 | 90.4 | 77.6 | 68.8 | 60.8 | 74.4 | |
| Grid searchType=Static, Backbone=Qwen3-4B, Iterations=162026.06 | 87.2 | 84 | 69.6 | 69.6 | 77.6 | |
| PractitionerType=Static, Backbone=Qwen3-4B, Iterations=162026.06 | 76.8 | 60 | 65.6 | 60.8 | 65.8 | |
| Qwen3-4BType=–, Backbone=Qwen3-4B, Iterations=162026.06 | 60 | 44.8 | 50.4 | 50.4 | 51.4 |