alon-albalak/qwen-4b-noveltybench-comprehensive-evaluation

--- dataset_info: features: - name: id dtype: string - name: prompt dtype: string - name: messages list: - name: content dtype: string - name: role dtype: string - name: completions list: string - name: partition list: int64 - name: distinct dtype: int64 - name: mean_intra_distance dtype: float64 - name: std_intra_distance dtype: float64 - name: min_intra_distance dtype: float64 - name: max_intra_distance dtype: float64 - name: num_pairs dtype: int64 - name: group_size dtype: int64 - name: reward_scores list: float64 - name: mean_reward dtype: float64 - name: std_reward dtype: float64 - name: min_reward dtype: float64 - name: max_reward dtype: float64 - name: judge_responses list: string - name: judge_thinking list: string - name: judge_scores list: int64 - name: mean_judge_score dtype: float64 - name: std_judge_score dtype: float64 - name: min_judge_score dtype: float64 - name: max_judge_score dtype: float64 - name: valid_scores_count dtype: int64 - name: total_completions dtype: int64 - name: completion_scores list: int64 - name: partition_scores list: int64 - name: utility dtype: float64 splits: - name: train num_bytes: 4950504 num_examples: 100 download_size: 1655696 dataset_size: 4950504 configs: - config_name: default data_files: - split: train path: data/train-* ---