Prompt Tool Sensitivity - Music Supplied Report: leaderboard
Prompt Tool Sensitivity: matched Music tasks, equally weighted over five wordings and written forms; unparsed answers fail. vLLM 0.31 uses greedy decoding with Qwen3 thinking disabled. Exact Hugging Face checkpoints and precision follow Table S1. Plain-text tool loop and canonical tool inputs; conditions remain separate. Supplied oracle answers and constant tools are explicit controls. Accuracy (0–1).
Metric: Accuracy (0–1). Source: arxiv.org. Saturation forecast: Estimated already saturated. 6 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Llama 3.1 8B Instruct | 0.59 | #1044 |
| 2 | Qwen 3 8B (Non-reasoning) | 0.59 | #681 (Qwen 3 8B) |
| 3 | Gemma 3 12B (IT) | 0.57 | #670 |
No result here: #5 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=prompt-tool-sensitivity-music-supplied-report · How It Works · Data refreshed daily, snapshot 2026-10-25.