{ "total_groups": 5, "rs_cross_stats": { "mean": 0.1759047640612712, "std": 0.14271632091758432, "min": 0.05101916888623692, "max": 0.4316129517679338 }, "groups_sorted_by_rs_cross": [ { "name": "group_0_(最短)", "rs_cross": 0.4316129517679338, "num_texts": 10, "avg_length": 329.6, "r_effs": { "llama3.2-3b-instruct": 2.226330041885376, "Mistral-7B-v0.3": 2.4710209369659424, "llama3-8b": 2.1663646697998047, "Qwen2.5-7B-Instruct": 3.8829245567321777, "gemma-2-9b-it": 2.18928599357605 } }, { "name": "group_1_(较短)", "rs_cross": 0.23267617354590583, "num_texts": 10, "avg_length": 446.0, "r_effs": { "llama3.2-3b-instruct": 2.1320927143096924, "Mistral-7B-v0.3": 2.282879590988159, "llama3-8b": 1.977645993232727, "Qwen2.5-7B-Instruct": 3.2836060523986816, "gemma-2-9b-it": 2.0314362049102783 } }, { "name": "group_2_(中等)", "rs_cross": 0.08646117594966882, "num_texts": 10, "avg_length": 555.8, "r_effs": { "llama3.2-3b-instruct": 2.0396993160247803, "Mistral-7B-v0.3": 2.237123727798462, "llama3-8b": 1.899621844291687, "Qwen2.5-7B-Instruct": 2.7096762657165527, "gemma-2-9b-it": 1.9543957710266113 } }, { "name": "group_4_(最长)", "rs_cross": 0.0777543501566106, "num_texts": 10, "avg_length": 859.0, "r_effs": { "llama3.2-3b-instruct": 2.1292803287506104, "Mistral-7B-v0.3": 2.7304728031158447, "llama3-8b": 2.1836905479431152, "Qwen2.5-7B-Instruct": 2.7516024112701416, "gemma-2-9b-it": 2.210625410079956 } }, { "name": "group_3_(较长)", "rs_cross": 0.05101916888623692, "num_texts": 10, "avg_length": 684.8, "r_effs": { "llama3.2-3b-instruct": 2.1218972206115723, "Mistral-7B-v0.3": 2.568671703338623, "llama3-8b": 2.1244308948516846, "Qwen2.5-7B-Instruct": 2.597262144088745, "gemma-2-9b-it": 2.120511054992676 } } ], "correlations": { "rs_cross_vs_text_length": { "spearman_rho": -0.8999999999999998, "spearman_pvalue": 0.03738607346849874, "n_samples": 5, "interpretation": "正相关表示长文本(复杂)导致更高 RS_cross" }, "rs_cross_vs_estimated_difficulty": { "spearman_rho": 0.49999999999999994, "spearman_pvalue": 0.3910022189557703, "n_samples": 5, "interpretation": "正相关表示高难度导致更高 RS_cross" }, "rs_cross_trend": { "spearman_rho": -0.8999999999999998, "spearman_pvalue": 0.03738607346849874, "interpretation": "正相关表示 RS_cross 随文本长度增加而上升" } } }