kth8/Qwen3.5-4B-SuperGPQA-benchmark

--- license: apache-2.0 language: - en base_model: Qwen/Qwen3.5-4B datasets: - m-a-p/SuperGPQA --- Benchmark of [Qwen/Qwen3.5-4B](https://huggingface.co/Qwen/Qwen3.5-4B) against [m-a-p/SuperGPQA](https://huggingface.co/datasets/m-a-p/SuperGPQA) dataset. Accuracy: 47.8% with Python tool. | Metric | Value | |----------------------|---------------| | **Correct** | 478 | | **Incorrect** | 522 | | **Errors** | 0 | | **Total samples** | 1000 | | **Python tool calls**| 1666 | | **Total completion tokens** | 3,977,352 | Raw stats: ```json { "accuracy": 0.478, "correct": 478, "incorrect": 522, "error": 0, "total": 1000, "python_tool_calls": 1666, "completion_tokens": 3977352 } ```