Judgemark-v2lp/results/scores.csv

1	model	judgemark_score	judgemark_score_raw	kendall_tau_bootstrapped	std_dev	kw_stat	ci99_overlap_magnitude_sum_norm	calibrated_score_range_norm	modulated_ci95	emd_norm
2	ministral/Ministral-3b-instruct	7.5	15.06	0.0	0.099	0.013	0.094	0.858	0.012	0.363
3	gpt-4o-mini	44.05	44.35	0.628	0.438	0.298	0.324	3.149	0.109	1.187
4	deepseek-ai/deepseek-r1	76.97	71.77	0.863	0.821	0.905	0.683	6.098	0.6	2.195
5	claude-3.5-haiku-20241022	77.64	76.88	0.898	0.832	0.891	0.665	5.975	0.575	2.229
6	claude-3.5-sonnet-20240620	78.84	74.72	0.905	0.83	0.902	0.664	6.255	0.591	2.232
7	microsoft/wizardlm-2-8x22b	55.42	52.86	0.749	0.566	0.463	0.417	4.62	0.198	1.53
8	gpt-4o-2024-11-20	82.51	77.75	0.89	0.837	0.985	0.75	6.233	0.814	2.228
9	gemini-2.0-flash-thinking-exp-1219	71.79	66.1	0.818	0.765	0.817	0.631	5.519	0.453	2.057
10	deepseek-ai/deepseek-r1-distill-llama-70b	59.67	56.86	0.764	0.642	0.6	0.474	4.485	0.243	1.726
11	mistralai/mistral-large-instruct-2411	65.63	63.25	0.794	0.732	0.636	0.505	5.873	0.296	1.935
12	meta-llama/llama-3.1-8b-instruct	35.72	35.67	0.46	0.401	0.201	0.207	3.232	0.063	1.075
13	meta-llama/llama-3.1-405b-instruct	60.09	58.76	0.791	0.633	0.582	0.471	4.515	0.243	1.712
14	deepseek/deepseek-r1-distill-qwen-14b
15	mistralai/mistral-small-24b-instruct-2501	63.64	61.23	0.808	0.67	0.618	0.504	5.194	0.286	1.802
16	liquid/lfm-7b	10.24	14.33	0.139	0.192	0.027	0.0	1.702	0.0	0.584