Datasets:

Indigma
/

SeLMRoute-Semantic-Features

sample_id string	det_char_count int64	det_word_count int64	det_line_count int64	det_question_marks int64	det_digit_count int64	det_mean_word_len float64	jev_ambiguity__confidence float64	jev_ambiguity__entropy float64	jev_ambiguity__p__0 float64	jev_ambiguity__p__1 float64	jev_ambiguity__p__2 float64	jev_ambiguity__p__3 float64	jev_ambiguity__score float64	jev_ambiguity__expected float64	jev_ambiguity__spread float64	jev_answer_openness__confidence float64	jev_answer_openness__entropy float64	jev_answer_openness__p__0 float64	jev_answer_openness__p__1 float64	jev_answer_openness__p__2 float64	jev_answer_openness__p__3 float64	jev_answer_openness__score float64	jev_answer_openness__expected float64	jev_answer_openness__spread float64	jev_code_reasoning__noul float64	jev_code_reasoning__entropy float64	jev_constraint_density__confidence float64	jev_constraint_density__entropy float64	jev_constraint_density__p__0 float64	jev_constraint_density__p__1 float64	jev_constraint_density__p__2 float64	jev_constraint_density__p__3 float64	jev_constraint_density__score float64	jev_constraint_density__expected float64	jev_constraint_density__spread float64	jev_context_integration__confidence float64	jev_context_integration__entropy float64	jev_context_integration__p__0 float64	jev_context_integration__p__1 float64	jev_context_integration__p__2 float64	jev_context_integration__p__3 float64	jev_context_integration__score float64	jev_context_integration__expected float64	jev_context_integration__spread float64	jev_current_information__noul float64	jev_current_information__entropy float64	jev_decomposition_need__confidence float64	jev_decomposition_need__entropy float64	jev_decomposition_need__p__0 float64	jev_decomposition_need__p__1 float64	jev_decomposition_need__p__2 float64	jev_decomposition_need__p__3 float64	jev_decomposition_need__score float64	jev_decomposition_need__expected float64	jev_decomposition_need__spread float64	jev_domain_specialization__confidence float64	jev_domain_specialization__entropy float64	jev_domain_specialization__p__0 float64	jev_domain_specialization__p__1 float64	jev_domain_specialization__p__2 float64	jev_domain_specialization__p__3 float64	jev_domain_specialization__score float64	jev_domain_specialization__expected float64	jev_domain_specialization__spread float64	jev_exactness__confidence float64	jev_exactness__entropy float64	jev_exactness__p__2 float64	jev_exactness__p__3 float64	jev_exactness__score float64	jev_exactness__expected float64	jev_exactness__spread float64	jev_external_knowledge__noul float64	jev_external_knowledge__entropy float64	jev_factual_recall__noul float64	jev_factual_recall__entropy float64	jev_formal_logic__noul float64	jev_formal_logic__entropy float64	jev_math_reasoning__noul float64	jev_math_reasoning__entropy float64	jev_reasoning_depth__confidence float64	jev_reasoning_depth__entropy float64	jev_reasoning_depth__p__0 float64	jev_reasoning_depth__p__1 float64	jev_reasoning_depth__p__2 float64	jev_reasoning_depth__p__3 float64	jev_reasoning_depth__score float64	jev_reasoning_depth__expected float64	jev_reasoning_depth__spread float64	jev_social_affective__noul float64	jev_social_affective__entropy float64	jev_task_family__confidence float64	jev_task_family__entropy float64	jev_task_family__p__code float64	jev_task_family__p__knowledge float64	jev_task_family__p__logic float64	jev_task_family__p__mathematics float64	jev_tool_interaction__noul float64	jev_tool_interaction__entropy float64	meta_jev_input_tokens float64	meta_jev_latency_ms float64
aime:hybrid:10:7d2b9ecfeeda	237	44	1	0	4	4.409091	0.46	0.691733	0.64	0.24	0.08	0.04	0.54	0.52	0.805978	0.86	0.24109	0.92	0.04	0.04	0	0.14	0.12	0.430813	0.08	0.402179	0.2	0.807683	0.41	0.39	0.18	0.02	0.8	0.81	0.796178	0.53	0.709968	0.01	0.37	0.52	0.1	1.72	1.71	0.65261	0.02	0.141441	0.6	0.610028	0.03	0.33	0.62	0.02	1.65	1.63	0.577148	0.64	0.591134	0.04	0.26	0.68	0.02	1.68	1.68	0.581034	1	-0	0	1	3	3	0	0.15	0.60984	0.07	0.365924	0.94	0.327445	0.98	0.141441	0.77	0.489427	0	0.1	0.78	0.12	2.01	2.02	0.468615	0.02	0.141441	0.66	0.293764	0	0	0.3	0.7	0.12	0.529361	1,512	811.247546
aime:hybrid:11:f2fe46bf765d	117	18	1	0	11	5.555556	0.95	0.097196	0.97	0.03	0	0	0.05	0.03	0.170587	1	-0	1	0	0	0	0	0	0	0.09	0.43647	0.42	0.777721	0.4	0.42	0.17	0.01	0.79	0.79	0.752263	0.46	0.651841	0.01	0.5	0.44	0.05	1.51	1.53	0.607536	0.02	0.141441	0.48	0.682205	0.07	0.48	0.44	0.01	1.39	1.39	0.630793	0.87	0.318979	0	0.11	0.87	0.02	1.9	1.91	0.349142	0.99	-0	0	1	2.99	3	0	0.27	0.841465	0.07	0.365924	0.42	0.981454	0.99	0.080793	0.78	0.499742	0.01	0.1	0.79	0.1	1.98	1.98	0.48949	0.01	0.080793	1	-0	0	0	0	1	0.15	0.60984	1,499	762.556063
aime:hybrid:12:73df31e8148e	394	78	3	0	26	4.051282	0.17	0.841912	0.41	0.39	0.15	0.05	0.83	0.84	0.856971	0.96	0.08072	0.98	0.01	0.01	0	0.04	0.03	0.221585	0.1	0.468996	0.48	0.648142	0	0.58	0.33	0.09	1.52	1.51	0.655668	0.55	0.697639	0	0.14	0.55	0.31	2.18	2.17	0.648922	0.02	0.141441	0.66	0.620173	0.03	0.24	0.68	0.05	1.75	1.75	0.589491	0.52	0.724479	0.17	0.52	0.31	0	1.14	1.14	0.678528	1	-0	0	1	3	3	0	0.19	0.701471	0.05	0.286397	0.95	0.286397	0.98	0.141441	0.72	0.488672	0	0.03	0.73	0.24	2.21	2.21	0.475289	0.02	0.141441	0.45	0.332949	0	0	0.52	0.48	0.09	0.43647	1,591	764.305205
aime:hybrid:13:60f7c1749379	519	92	2	0	8	4.652174	0.8	0.386626	0.84	0.13	0.02	0.01	0.2	0.2	0.509902	1	-0	1	0	0	0	0	0	0	0.06	0.327445	0.02	0.845028	0.35	0.34	0.29	0.02	0.98	0.98	0.848292	0.67	0.61972	0	0.14	0.67	0.19	2.05	2.05	0.572276	0.02	0.141441	0.76	0.418678	0.01	0.22	0.77	0	1.77	1.76	0.449889	0.59	0.565668	0.38	0.59	0.03	0	0.66	0.65	0.53619	1	-0	0	1	3	3	0	0.06	0.327445	0.03	0.194392	0.3	0.881291	0.99	0.080793	0.9	0.252624	0	0.02	0.91	0.07	2.05	2.05	0.295804	0.02	0.141441	1	-0	0	0	0	1	0.07	0.365924	1,583	777.22836
aime:hybrid:14:10c80ba48d18	266	46	1	0	6	4.804348	0.43	0.735722	0.62	0.23	0.09	0.06	0.57	0.59	0.884251	1	-0	1	0	0	0	0	0	0	0.16	0.63431	0.39	0.870857	0.06	0.21	0.46	0.27	1.94	1.94	0.846404	0.5	0.731205	0	0.17	0.5	0.33	2.16	2.16	0.688767	0.02	0.141441	0.8	0.479802	0.01	0.07	0.8	0.12	2.04	2.03	0.478644	0.77	0.457068	0.02	0.19	0.78	0.01	1.79	1.78	0.481248	1	-0	0	1	3	3	0	0.25	0.811278	0.04	0.242292	0.87	0.557438	0.99	0.080793	0.51	0.535069	0	0.01	0.51	0.48	2.47	2.47	0.518748	0.01	0.080793	0.93	0.109148	0	0	0.06	0.94	0.21	0.741483	1,520	751.979978
aime:hybrid:15:c0aff28b8c4a	236	43	1	0	4	4.511628	0.43	0.846145	0.16	0.51	0.26	0.07	1.25	1.24	0.801499	0.99	-0	1	0	0	0	0.01	0	0	0.08	0.402179	0.36	0.80831	0.12	0.34	0.49	0.05	1.47	1.47	0.767529	0.43	0.6677	0.01	0.49	0.44	0.06	1.54	1.55	0.622495	0.02	0.141441	0.72	0.53833	0.02	0.22	0.73	0.03	1.78	1.77	0.526403	0.66	0.506157	0.02	0.3	0.68	0	1.66	1.66	0.514198	1	-0	0	1	3	3	0	0.35	0.934068	0.11	0.499916	0.96	0.242292	0.99	0.080793	0.8	0.436211	0	0.06	0.81	0.13	2.07	2.07	0.430232	0.02	0.141441	0.43	0.333237	0	0	0.49	0.51	0.08	0.402179	1,504	741.517406
aime:hybrid:16:4827451ee78b	205	35	1	0	6	4.885714	0.82	0.37064	0.84	0.14	0.01	0.01	0.18	0.19	0.483632	0.96	0.121146	0.96	0.04	0	0	0.04	0.04	0.195959	0.14	0.584239	0.43	0.764004	0.1	0.45	0.42	0.03	1.38	1.38	0.703989	0.49	0.643353	0.01	0.55	0.39	0.05	1.48	1.48	0.607947	0.03	0.194392	0.81	0.468996	0.01	0.09	0.81	0.09	1.98	1.98	0.468615	0.79	0.360964	0	0	0.8	0.2	2.2	2.2	0.4	1	-0	0	1	3	3	0	0.89	0.499916	0.2	0.721928	0.82	0.680077	0.98	0.141441	0.52	0.614964	0.01	0.03	0.52	0.44	2.39	2.39	0.598247	0.01	0.080793	0.99	0.026931	0	0.01	0	0.99	0.29	0.868721	1,513	772.469825
aime:hybrid:17:951370224398	339	58	2	0	5	4.862069	0.56	0.638269	0.67	0.24	0.06	0.03	0.44	0.45	0.739932	1	-0	1	0	0	0	0	0	0	0.08	0.402179	0.26	0.937344	0.14	0.28	0.4	0.17	1.6	1.59	0.939095	0.54	0.71159	0	0.17	0.55	0.28	2.1	2.11	0.66174	0.02	0.141441	0.79	0.499019	0.01	0.09	0.79	0.11	2.01	2	0.489898	0.81	0.422685	0	0.12	0.82	0.06	1.94	1.94	0.42	1	-0	0	1	3	3	0	0.21	0.741483	0.05	0.286397	0.73	0.841465	0.99	0.080793	0.49	0.560131	0	0.02	0.47	0.51	2.49	2.49	0.538424	0.01	0.080793	1	-0	0	0	0	1	0.24	0.79504	1,547	755.411629
aime:hybrid:18:14b236690314	228	40	1	0	2	4.725	0.43	0.851713	0.24	0.51	0.18	0.07	1.09	1.08	0.832827	0.98	0.040397	0.99	0.01	0	0	0.02	0.01	0.099499	0.15	0.60984	0.31	0.911304	0.21	0.38	0.33	0.08	1.26	1.28	0.884081	0.18	0.922057	0.09	0.39	0.3	0.22	1.64	1.65	0.920598	0.03	0.194392	0.68	0.615237	0.02	0.17	0.71	0.1	1.89	1.89	0.581292	0.64	0.582984	0.03	0.28	0.67	0.02	1.68	1.68	0.56356	1	-0	0	1	3	3	0	0.48	0.998846	0.17	0.657705	0.94	0.327445	0.98	0.141441	0.62	0.554728	0	0.03	0.62	0.35	2.32	2.32	0.526878	0.02	0.141441	0.45	0.332949	0	0	0.52	0.48	0.13	0.557438	1,503	823.561172
aime:hybrid:19:357603fe2192	154	19	1	0	3	7.157895	0.9	0.222832	0.92	0.07	0.01	0	0.1	0.09	0.319218	0.99	-0	1	0	0	0	0.01	0	0	0.05	0.286397	0.17	0.955154	0.24	0.2	0.4	0.16	1.48	1.48	1.0245	0.61	0.643949	0	0.11	0.62	0.27	2.16	2.16	0.595315	0.02	0.141441	0.74	0.532233	0.02	0.17	0.76	0.05	1.85	1.84	0.523832	0.7	0.440645	0	0	0.7	0.3	2.29	2.3	0.458258	1	-0	0	1	3	3	0	0.47	0.997402	0.11	0.499916	0.73	0.841465	0.99	0.080793	0.58	0.568686	0	0.03	0.58	0.39	2.35	2.36	0.538888	0.01	0.080793	1	-0	0	0	0	1	0.1	0.468996	1,507	763.081682
aime:hybrid:1:14aaf039bf2e	380	54	5	0	13	6.037037	0.97	0.097196	0.97	0.03	0	0	0.03	0.03	0.170587	1	-0	1	0	0	0	0	0	0	0.05	0.286397	0.24	0.93181	0.18	0.27	0.42	0.13	1.51	1.5	0.932738	0.61	0.659417	0	0.14	0.62	0.24	2.1	2.1	0.608276	0.02	0.141441	0.73	0.485584	0.02	0.22	0.75	0.01	1.75	1.75	0.497494	0.58	0.519207	0.01	0.39	0.6	0	1.59	1.59	0.511762	1	-0	0	1	3	3	0	0.13	0.557438	0.04	0.242292	0.73	0.841465	0.99	0.080793	0.84	0.36064	0	0.02	0.84	0.14	2.11	2.12	0.381576	0.01	0.080793	1	-0	0	0	0	1	0.08	0.402179	1,596	749.523053
aime:hybrid:20:dc0068844c4e	421	77	5	0	8	4.467532	0.12	0.867421	0.45	0.33	0.14	0.08	0.88	0.85	0.942072	0.99	-0	1	0	0	0	0.01	0	0	0.12	0.529361	0.57	0.707824	0	0.22	0.57	0.21	1.98	1.99	0.655668	0.53	0.6044	0	0.05	0.38	0.57	2.53	2.52	0.59127	0.02	0.141441	0.65	0.648367	0.01	0.1	0.65	0.24	2.12	2.12	0.604649	0.86	0.382828	0.01	0.07	0.86	0.06	1.98	1.97	0.411218	1	-0	0	1	3	3	0	0.14	0.584239	0.05	0.286397	0.97	0.194392	0.99	0.080793	0.49	0.581861	0	0.03	0.5	0.47	2.43	2.44	0.553534	0.01	0.080793	0.76	0.247161	0	0	0.79	0.21	0.15	0.60984	1,555	765.885861
aime:hybrid:21:a6581367f5e8	368	56	1	0	16	5.589286	0.86	0.289509	0.88	0.11	0.01	0	0.14	0.13	0.364829	1	-0	1	0	0	0	0	0	0	0.3	0.881291	0.3	0.894477	0.12	0.34	0.42	0.12	1.55	1.54	0.853464	0.49	0.753085	0	0.23	0.49	0.28	2.05	2.05	0.71239	0.02	0.141441	0.8	0.454371	0	0.07	0.8	0.13	2.05	2.06	0.44317	0.67	0.592695	0.03	0.23	0.7	0.04	1.75	1.75	0.572276	1	-0	0	1	3	3	0	0.27	0.841465	0.05	0.286397	0.89	0.499916	0.99	0.080793	0.56	0.527168	0	0.01	0.42	0.57	2.56	2.56	0.51614	0.01	0.080793	0.96	0.07398	0.01	0	0.02	0.97	0.45	0.992774	1,577	801.451855
aime:hybrid:22:01f4378fc0aa	233	42	1	0	9	4.571429	0.58	0.588625	0.65	0.3	0.04	0.01	0.42	0.41	0.617981	0.99	-0	1	0	0	0	0.01	0	0	0.06	0.327445	0.16	0.960528	0.23	0.22	0.39	0.16	1.48	1.48	1.014692	0.59	0.643902	0	0.09	0.59	0.32	2.23	2.23	0.597578	0.02	0.141441	0.65	0.582984	0.02	0.28	0.67	0.03	1.71	1.71	0.553082	0.84	0.386626	0.01	0.13	0.84	0.02	1.88	1.87	0.416053	1	-0	0	1	3	3	0	0.24	0.79504	0.07	0.365924	0.54	0.995378	0.99	0.080793	0.69	0.517683	0	0.03	0.69	0.28	2.24	2.25	0.497494	0.01	0.080793	1	-0	0	0	0	1	0.11	0.499916	1,513	741.174293
aime:hybrid:23:7166d6915f46	237	41	1	0	8	4.804878	0.5	0.673703	0.6	0.31	0.07	0.02	0.5	0.51	0.714073	0.99	-0	1	0	0	0	0.01	0	0	0.08	0.402179	0.29	0.923049	0.16	0.24	0.45	0.15	1.59	1.59	0.928386	0.49	0.71606	0	0.14	0.49	0.37	2.24	2.23	0.676092	0.02	0.141441	0.77	0.500056	0.01	0.15	0.78	0.06	1.88	1.89	0.48775	0.89	0.257948	0	0.01	0.9	0.09	2.08	2.08	0.305941	1	-0	0	1	3	3	0	0.32	0.904381	0.06	0.327445	0.75	0.811278	0.99	0.080793	0.57	0.524815	0	0.01	0.58	0.41	2.4	2.4	0.509902	0.01	0.080793	1	-0	0	0	0	1	0.15	0.60984	1,534	780.377336
aime:hybrid:24:c48bbfcdb9d7	303	60	9	0	3	4	0.33	0.912091	0.26	0.41	0.25	0.08	1.16	1.15	0.89861	0.82	0.322678	0.89	0.06	0.04	0.01	0.18	0.17	0.530189	0.09	0.43647	0.69	0.586743	0.01	0.06	0.69	0.24	2.16	2.16	0.560714	0.56	0.529222	0	0.01	0.56	0.43	2.41	2.42	0.51342	0.02	0.141441	0.84	0.406886	0.01	0.11	0.84	0.04	1.91	1.91	0.426497	0.63	0.647718	0.21	0.64	0.15	0	0.95	0.94	0.596992	1	-0	0	1	3	3	0	0.13	0.557438	0.04	0.242292	0.89	0.499916	0.99	0.080793	0.74	0.443948	0	0.01	0.74	0.25	2.24	2.24	0.449889	0.02	0.141441	0.66	0.289574	0	0	0.71	0.29	0.09	0.43647	1,528	776.414047
aime:hybrid:25:d7c47fba84b4	401	77	1	0	21	4.220779	0.54	0.645056	0.71	0.15	0.09	0.05	0.46	0.48	0.854166	0.99	-0	1	0	0	0	0.01	0	0	0.07	0.365924	0.34	0.902771	0.16	0.16	0.49	0.19	1.73	1.71	0.951788	0.72	0.451585	0	0.01	0.26	0.73	2.72	2.72	0.470744	0.02	0.141441	0.51	0.651419	0.06	0.36	0.57	0.01	1.54	1.53	0.623779	0.59	0.692458	0.22	0.59	0.19	0	0.98	0.97	0.639609	1	-0	0	1	3	3	0	0.1	0.468996	0.03	0.194392	0.93	0.365924	0.99	0.080793	0.76	0.542349	0.01	0.12	0.76	0.11	1.98	1.97	0.518748	0.02	0.141441	0.44	0.332949	0	0	0.48	0.52	0.05	0.286397	1,564	276.528658
aime:hybrid:26:869c98209496	830	129	3	0	36	5.364341	0.97	0.07072	0.98	0.02	0	0	0.03	0.02	0.14	1	-0	1	0	0	0	0	0	0	0.06	0.327445	0.22	0.950275	0.3	0.34	0.25	0.11	1.18	1.17	0.980357	0.56	0.594188	0	0.05	0.35	0.6	2.56	2.55	0.589491	0.02	0.141441	0.8	0.468192	0.01	0.08	0.81	0.1	2.01	2	0.469042	0.83	0.328852	0	0	0.83	0.17	2.16	2.17	0.375633	1	-0	0	1	3	3	0	0.22	0.760168	0.07	0.365924	0.75	0.811278	0.99	0.080793	0.56	0.492908	0	0	0.43	0.57	2.56	2.57	0.495076	0.02	0.141441	1	-0	0	0	0	1	0.16	0.63431	1,807	335.908234
aime:hybrid:27:57c667dbfc4c	564	107	1	0	4	4.280374	0.64	0.568558	0.71	0.23	0.04	0.02	0.36	0.37	0.658103	1	-0	1	0	0	0	0	0	0	0.05	0.286397	0.39	0.866975	0.2	0.42	0.33	0.05	1.22	1.23	0.822861	0.53	0.613594	0	0.05	0.53	0.42	2.36	2.37	0.577148	0.02	0.141441	0.79	0.471935	0.01	0.16	0.79	0.04	1.87	1.86	0.469468	0.78	0.467022	0.01	0.2	0.77	0.02	1.81	1.8	0.469042	1	-0	0	1	3	3	0	0.17	0.657705	0.05	0.286397	0.35	0.934068	0.99	0.080793	0.67	0.489788	0	0.01	0.67	0.32	2.3	2.31	0.483632	0.01	0.080793	1	-0	0	0	0	1	0.1	0.468996	1,610	289.197522
aime:hybrid:28:d8c514307d5e	449	77	1	0	6	4.844156	0.96	0.097196	0.97	0.03	0	0	0.04	0.03	0.170587	1	-0	1	0	0	0	0	0	0	0.06	0.327445	0.4	0.837581	0.09	0.32	0.49	0.1	1.61	1.6	0.787401	0.55	0.579274	0	0.04	0.36	0.6	2.55	2.56	0.571314	0.02	0.141441	0.73	0.47216	0	0.02	0.73	0.25	2.23	2.23	0.46594	0.88	0.289509	0	0.01	0.88	0.11	2.1	2.1	0.331662	1	-0	0	1	3	3	0	0.34	0.924819	0.06	0.327445	0.5	1	0.99	0.080793	0.74	0.405639	0	0	0.25	0.75	2.74	2.75	0.433013	0.02	0.141441	1	-0	0	0	0	1	0.18	0.680077	1,601	278.838304
aime:hybrid:29:5650d2a0767a	181	28	1	1	10	5.5	0.34	0.91104	0.26	0.43	0.22	0.09	1.14	1.14	0.905759	0.99	-0	1	0	0	0	0.01	0	0	0.07	0.365924	0.32	0.907327	0.16	0.21	0.48	0.15	1.62	1.62	0.924986	0.71	0.458845	0	0.01	0.27	0.72	2.71	2.71	0.475289	0.02	0.141441	0.69	0.571009	0.03	0.23	0.71	0.03	1.75	1.74	0.558928	0.84	0.357408	0.01	0.13	0.85	0.01	1.87	1.86	0.4005	1	-0	0	1	3	3	0	0.22	0.760168	0.04	0.242292	0.59	0.9765	0.99	0.080793	0.63	0.504304	0	0.01	0.64	0.35	2.34	2.34	0.494368	0.01	0.080793	1	-0	0	0	0	1	0.14	0.584239	1,528	358.750414
aime:hybrid:2:1bf70e3660d6	511	81	1	0	9	5.320988	0.92	0.20367	0.93	0.06	0.01	0	0.08	0.08	0.305941	1	-0	1	0	0	0	0	0	0	0.07	0.365924	0.19	0.955008	0.2	0.26	0.39	0.15	1.49	1.49	0.974628	0.53	0.573796	0	0.03	0.41	0.56	2.53	2.53	0.555968	0.02	0.141441	0.76	0.503733	0.01	0.17	0.77	0.05	1.86	1.86	0.490306	0.83	0.382057	0	0.12	0.84	0.04	1.91	1.92	0.391918	1	-0	0	1	3	3	0	0.17	0.657705	0.06	0.327445	0.59	0.9765	0.99	0.080793	0.52	0.53361	0	0.01	0.46	0.53	2.52	2.52	0.51923	0.01	0.080793	1	-0	0	0	0	1	0.15	0.60984	1,628	326.560835
aime:hybrid:30:b0156a83da3b	141	25	2	0	13	4.64	0.68	0.526785	0.72	0.24	0.03	0.01	0.32	0.33	0.584038	1	-0	1	0	0	0	0	0	0	0.12	0.529361	0.72	0.479802	0.8	0.12	0.07	0.01	0.28	0.29	0.637103	0.46	0.780722	0.01	0.33	0.47	0.19	1.84	1.84	0.731027	0.02	0.141441	0.75	0.555324	0.01	0.11	0.75	0.13	2	2	0.52915	0.86	0.292119	0	0	0.86	0.14	2.14	2.14	0.346987	1	-0	0	1	3	3	0	0.55	0.992774	0.1	0.468996	0.73	0.841465	0.99	0.080793	0.54	0.557034	0	0.02	0.54	0.44	2.41	2.42	0.532541	0.01	0.080793	1	-0	0	0	0	1	0.21	0.741483	1,512	366.339581
aime:hybrid:31:a3145c72cb6b	84	16	1	0	5	4.3125	0.72	0.471554	0.78	0.18	0.03	0.01	0.28	0.27	0.563116	1	-0	1	0	0	0	0	0	0	0.12	0.529361	0.5	0.650765	0.05	0.51	0.43	0.01	1.41	1.4	0.6	0.51	0.636036	0.02	0.54	0.41	0.03	1.44	1.45	0.589491	0.02	0.141441	0.51	0.575893	0.03	0.42	0.55	0	1.52	1.52	0.556417	0.49	0.659906	0.06	0.39	0.54	0.01	1.5	1.5	0.6245	1	-0	0	1	3	3	0	0.29	0.868721	0.07	0.365924	0.72	0.855451	0.99	0.080793	0.81	0.422493	0	0.13	0.81	0.05	1.91	1.9	0.458258	0.01	0.080793	0.99	0.026931	0	0	0.01	0.99	0.11	0.499916	1,481	260.635107
aime:hybrid:32:9592e8215196	381	64	1	0	13	4.96875	0.36	0.77343	0.56	0.27	0.13	0.04	0.64	0.65	0.852936	0.99	-0	1	0	0	0	0.01	0	0	0.06	0.327445	0	0.927996	0.34	0.08	0.3	0.28	1.52	1.52	1.220492	0.93	0.182962	0	0	0.07	0.93	2.93	2.93	0.255147	0.02	0.141441	0.75	0.447178	0	0.02	0.76	0.22	2.2	2.2	0.447214	0.91	0.273333	0.01	0.06	0.91	0.02	1.95	1.94	0.341174	1	-0	0	1	3	3	0	0.23	0.778011	0.04	0.242292	0.44	0.989588	0.99	0.080793	0.63	0.471342	0	0	0.36	0.64	2.63	2.64	0.48	0.02	0.141441	1	-0	0	0	0	1	0.14	0.584239	1,592	262.398563
aime:hybrid:33:9cbaa17c63e1	516	90	1	0	5	4.744444	0.82	0.344138	0.84	0.15	0.01	0	0.18	0.17	0.401373	1	-0	1	0	0	0	0	0	0	0.15	0.60984	0.56	0.764105	0.05	0.21	0.6	0.14	1.83	1.83	0.721873	0.48	0.710355	0	0.13	0.48	0.39	2.26	2.26	0.672607	0.02	0.141441	0.83	0.382057	0	0.12	0.84	0.04	1.92	1.92	0.391918	0.51	0.591483	0.04	0.4	0.56	0	1.52	1.52	0.574108	1	-0	0	1	3	3	0	0.1	0.468996	0.03	0.194392	0.97	0.194392	0.99	0.080793	0.68	0.489788	0	0.01	0.67	0.32	2.31	2.31	0.483632	0.02	0.141441	0.5	0.329863	0	0	0.44	0.56	0.26	0.826746	1,562	331.552763
aime:hybrid:34:981c7f2ec58b	147	22	1	0	12	5.727273	0.88	0.289509	0.88	0.11	0.01	0	0.12	0.13	0.364829	1	-0	1	0	0	0	0	0	0	0.13	0.557438	0.45	0.635297	0.04	0.45	0.5	0.01	1.47	1.48	0.59127	0.49	0.59504	0.01	0.51	0.46	0.02	1.49	1.49	0.556687	0.02	0.141441	0.65	0.554659	0.02	0.32	0.65	0.01	1.66	1.65	0.53619	0.47	0.581266	0.03	0.46	0.51	0	1.48	1.48	0.556417	1	-0	0	1	3	3	0	0.23	0.778011	0.07	0.365924	0.85	0.60984	0.99	0.080793	0.86	0.352538	0	0.04	0.86	0.1	2.06	2.06	0.369324	0.01	0.080793	0.99	0.026931	0	0	0.01	0.99	0.13	0.557438	1,505	274.469589
aime:hybrid:35:7528e62e74ea	218	35	1	0	20	5.257143	0.92	0.20367	0.93	0.06	0.01	0	0.08	0.08	0.305941	1	-0	1	0	0	0	0	0	0	0.13	0.557438	0	0.944775	0.32	0.26	0.32	0.1	1.21	1.2	1	0.52	0.743112	0.01	0.31	0.53	0.15	1.83	1.82	0.683813	0.02	0.141441	0.8	0.482888	0.01	0.08	0.8	0.11	2.03	2.01	0.479479	0.61	0.578154	0.03	0.32	0.64	0.01	1.63	1.63	0.559553	1	-0	0	1	3	3	0	0.45	0.992774	0.05	0.286397	0.95	0.286397	0.99	0.080793	0.65	0.522334	0	0.02	0.65	0.33	2.3	2.31	0.503885	0.02	0.141441	0.78	0.226692	0	0	0.18	0.82	0.23	0.778011	1,523	317.480194
aime:hybrid:36:e93030b1d230	252	47	1	0	5	4.382979	0.95	0.121146	0.96	0.04	0	0	0.05	0.04	0.195959	0.99	-0	1	0	0	0	0.01	0	0	0.04	0.242292	0.4	0.827998	0.13	0.28	0.52	0.07	1.54	1.53	0.805667	0.63	0.65236	0	0.14	0.63	0.23	2.09	2.09	0.601581	0.02	0.141441	0.76	0.448022	0.01	0.21	0.77	0.01	1.78	1.78	0.46	0.83	0.317155	0	0.16	0.84	0	1.84	1.84	0.366606	1	-0	0	1	3	3	0	0.22	0.760168	0.08	0.402179	0.62	0.958042	0.99	0.080793	0.88	0.28758	0	0.02	0.89	0.09	2.07	2.07	0.324191	0.01	0.080793	1	-0	0	0	0	1	0.05	0.286397	1,530	306.294759
aime:hybrid:37:70629b6e7c5d	455	70	1	0	0	5.514286	0.88	0.257948	0.9	0.09	0.01	0	0.12	0.11	0.343366	1	-0	1	0	0	0	0	0	0	0.08	0.402179	0.26	0.935598	0.28	0.35	0.28	0.09	1.18	1.18	0.942125	0.47	0.741375	0	0.18	0.48	0.34	2.15	2.16	0.703136	0.02	0.141441	0.49	0.703419	0.05	0.36	0.55	0.04	1.58	1.58	0.650846	0.49	0.690042	0.11	0.5	0.39	0	1.29	1.28	0.649307	1	-0	0	1	3	3	0	0.12	0.529361	0.04	0.242292	0.96	0.242292	0.99	0.080793	0.67	0.629729	0.01	0.11	0.68	0.2	2.07	2.07	0.587452	0.02	0.141441	0.5	0.328605	0	0	0.43	0.57	0.08	0.402179	1,571	332.114004
aime:hybrid:38:b552420b5a18	281	44	1	0	8	5.409091	0.78	0.398283	0.84	0.12	0.03	0.01	0.22	0.21	0.534696	1	-0	1	0	0	0	0	0	0	0.08	0.402179	0.4	0.857397	0.11	0.28	0.5	0.11	1.61	1.61	0.823347	0.54	0.623852	0	0.06	0.55	0.39	2.33	2.33	0.584038	0.02	0.141441	0.84	0.359065	0	0.03	0.85	0.12	2.09	2.09	0.376696	0.93	0.217445	0	0.03	0.93	0.04	2	2.01	0.264386	1	-0	0	1	3	3	0	0.24	0.79504	0.04	0.242292	0.74	0.826746	0.99	0.080793	0.58	0.490727	0	0	0.42	0.58	2.58	2.58	0.493559	0.01	0.080793	1	-0	0	0	0	1	0.23	0.778011	1,549	359.488109
aime:hybrid:39:edc7cd228836	332	47	1	0	4	6.085106	0.91	0.222832	0.92	0.07	0.01	0	0.09	0.09	0.319218	1	-0	1	0	0	0	0	0	0	0.1	0.468996	0.42	0.863048	0.08	0.25	0.5	0.17	1.77	1.76	0.826075	0.51	0.60962	0	0.05	0.4	0.55	2.51	2.5	0.591608	0.02	0.141441	0.73	0.480446	0	0.03	0.74	0.23	2.19	2.2	0.469042	0.93	0.213171	0	0.02	0.93	0.05	2.04	2.03	0.262869	1	-0	0	1	3	3	0	0.23	0.778011	0.04	0.242292	0.57	0.985815	0.99	0.080793	0.78	0.370741	0	0	0.21	0.79	2.78	2.79	0.407308	0.01	0.080793	1	-0	0	0	0	1	0.33	0.914926	1,549	269.434909
aime:hybrid:3:84c96fbdf01d	463	79	1	0	9	4.873418	0.88	0.280594	0.9	0.08	0.01	0.01	0.12	0.13	0.439431	1	-0	1	0	0	0	0	0	0	0.05	0.286397	0.45	0.782996	0.15	0.47	0.36	0.02	1.24	1.25	0.726292	0.54	0.680215	0	0.35	0.54	0.11	1.75	1.76	0.63435	0.02	0.141441	0.78	0.446641	0.01	0.18	0.79	0.02	1.81	1.82	0.455631	0.42	0.618161	0.05	0.47	0.48	0	1.43	1.43	0.587452	1	-0	0	1	3	3	0	0.1	0.468996	0.04	0.242292	0.97	0.194392	0.99	0.080793	0.84	0.388838	0	0.05	0.84	0.11	2.06	2.06	0.395474	0.02	0.141441	0.73	0.265013	0	0	0.24	0.76	0.07	0.365924	1,573	321.316526
aime:hybrid:40:f48bd392bc6e	640	147	5	0	37	3.306122	0.69	0.509583	0.79	0.13	0.05	0.03	0.31	0.32	0.705408	0.99	-0	1	0	0	0	0.01	0	0	0.13	0.557438	0	0.965493	0.19	0.18	0.25	0.38	1.82	1.82	1.134725	0.47	0.693458	0.01	0.1	0.32	0.57	2.47	2.45	0.71239	0.02	0.141441	0.48	0.645806	0.02	0.04	0.37	0.57	2.48	2.49	0.670746	0.64	0.687224	0.02	0.17	0.65	0.16	1.96	1.95	0.638357	1	-0	0	1	3	3	0	0.55	0.992774	0.07	0.365924	0.96	0.242292	0.99	0.080793	0.72	0.485584	0.01	0.02	0.22	0.75	2.72	2.71	0.553082	0.02	0.141441	0.52	0.3255	0	0	0.59	0.41	0.22	0.760168	1,697	323.423866
aime:hybrid:41:1352bde57c17	583	84	1	0	10	5.952381	0.86	0.29758	0.89	0.09	0.01	0.01	0.14	0.14	0.447661	1	-0	1	0	0	0	0	0	0	0.08	0.402179	0.3	0.92052	0.15	0.22	0.46	0.17	1.65	1.65	0.931397	0.67	0.48429	0	0.01	0.31	0.68	2.67	2.67	0.491019	0.02	0.141441	0.76	0.447178	0	0.02	0.76	0.22	2.2	2.2	0.447214	0.86	0.341199	0	0.08	0.87	0.05	1.96	1.97	0.359305	1	-0	0	1	3	3	0	0.15	0.60984	0.03	0.194392	0.46	0.995378	0.99	0.080793	0.73	0.413373	0	0	0.26	0.74	2.73	2.74	0.438634	0.02	0.141441	1	-0	0	0	0	1	0.27	0.841465	1,648	347.087713
aime:hybrid:42:3f8dfc841f22	405	66	1	0	3	5.151515	0.89	0.222832	0.92	0.07	0.01	0	0.11	0.09	0.319218	1	-0	1	0	0	0	0	0	0	0.1	0.468996	0.11	0.989511	0.19	0.25	0.31	0.25	1.61	1.62	1.05622	0.71	0.458845	0	0.01	0.27	0.72	2.71	2.71	0.475289	0.02	0.141441	0.59	0.485475	0	0	0.4	0.6	2.59	2.6	0.489898	0.74	0.443948	0	0.01	0.74	0.25	2.24	2.24	0.449889	1	-0	0	1	3	3	0	0.31	0.893173	0.04	0.242292	0.77	0.778011	0.99	0.080793	0.9	0.234498	0	0	0.1	0.9	2.9	2.9	0.3	0.02	0.141441	1	-0	0	0	0	1	0.31	0.893173	1,559	294.922313
aime:hybrid:43:12f6b189436f	388	65	1	0	4	4.984615	0.43	0.844071	0.27	0.5	0.17	0.06	1.03	1.02	0.824379	0.99	-0	1	0	0	0	0.01	0	0	0.09	0.43647	0.41	0.666358	0.69	0.09	0.17	0.05	0.59	0.58	0.94	0.54	0.757361	0.01	0.22	0.54	0.23	2	1.99	0.699929	0.02	0.141441	0.67	0.646598	0.04	0.17	0.7	0.09	1.84	1.84	0.628013	0.79	0.454371	0	0.13	0.8	0.07	1.94	1.94	0.44317	0.99	0.040397	0.01	0.99	2.99	2.99	0.099499	0.61	0.9648	0.1	0.468996	0.9	0.468996	0.99	0.080793	0.59	0.626944	0.01	0.05	0.6	0.34	2.28	2.27	0.597578	0.02	0.141441	0.89	0.156332	0	0	0.1	0.9	0.14	0.584239	1,530	332.089265
aime:hybrid:44:9bfdf6197204	347	56	1	0	11	5.214286	0.74	0.453423	0.8	0.14	0.04	0.01	0.26	0.25	0.572276	1	-0	1	0	0	0	0	0	0	0.07	0.365924	0.16	0.97771	0.18	0.29	0.33	0.2	1.56	1.55	1.003743	0.81	0.378996	0	0.01	0.18	0.81	2.81	2.8	0.424264	0.02	0.141441	0.59	0.542755	0	0.02	0.6	0.38	2.35	2.36	0.52	0.7	0.440645	0	0	0.7	0.3	2.29	2.3	0.458258	1	-0	0	1	3	3	0	0.37	0.950672	0.05	0.286397	0.58	0.981454	0.99	0.080793	0.86	0.292119	0	0	0.14	0.86	2.86	2.86	0.346987	0.02	0.141441	1	-0	0	0	0	1	0.24	0.79504	1,576	324.14724
aime:hybrid:45:cc2ed8965f94	192	31	1	0	11	5.225806	0.93	0.183223	0.94	0.05	0.01	0	0.07	0.07	0.291719	1	-0	1	0	0	0	0	0	0	0.32	0.904381	0	0.974472	0.33	0.26	0.26	0.15	1.22	1.23	1.066349	0.39	0.77676	0	0.36	0.4	0.24	1.87	1.88	0.765245	0.02	0.141441	0.56	0.548728	0	0.02	0.4	0.58	2.56	2.56	0.535164	0.56	0.492908	0	0	0.57	0.43	2.43	2.43	0.495076	1	-0	0	1	3	3	0	0.57	0.985815	0.05	0.286397	0.94	0.327445	0.99	0.080793	0.83	0.30492	0	0	0.15	0.85	2.83	2.85	0.357071	0.01	0.080793	0.97	0.064797	0	0	0.03	0.97	0.43	0.985815	1,518	347.501617
aime:hybrid:46:aa6399130b64	263	79	1	0	14	2.341772	0.5	0.688974	0.62	0.26	0.1	0.02	0.5	0.52	0.754718	0.97	0.08072	0.98	0.01	0.01	0	0.03	0.03	0.221585	0.07	0.365924	0	0.928673	0.31	0.08	0.28	0.33	1.63	1.63	1.230081	0.86	0.278719	0	0	0.13	0.87	2.86	2.87	0.336303	0.02	0.141441	0.83	0.382057	0	0.04	0.84	0.12	2.08	2.08	0.391918	0.66	0.517683	0.03	0.28	0.69	0	1.67	1.66	0.533292	1	-0	0	1	3	3	0	0.21	0.741483	0.04	0.242292	0.48	0.998846	0.99	0.080793	0.58	0.524815	0	0.01	0.58	0.41	2.4	2.4	0.509902	0.01	0.080793	1	-0	0	0	0	1	0.18	0.680077	1,551	423.869369
aime:hybrid:47:ca1a4c4ffd58	106	27	1	0	5	2.962963	0.82	0.352827	0.87	0.09	0.03	0.01	0.18	0.18	0.517301	1	-0	1	0	0	0	0	0	0	0.13	0.557438	0.51	0.765711	0.27	0.52	0.2	0.01	0.94	0.95	0.71239	0.43	0.861716	0.26	0.49	0.18	0.07	1.06	1.06	0.846404	0.02	0.141441	0.72	0.529463	0.02	0.21	0.74	0.03	1.77	1.78	0.521153	0.76	0.457068	0.02	0.19	0.78	0.01	1.78	1.78	0.481248	1	-0	0	1	3	3	0	0.23	0.778011	0.05	0.286397	0.8	0.721928	0.99	0.080793	0.7	0.486858	0	0.02	0.71	0.27	2.25	2.25	0.47697	0.01	0.080793	1	-0	0	0	0	1	0.2	0.721928	1,487	359.591159
aime:hybrid:48:951031f375bc	245	50	1	0	6	3.92	0.65	0.560218	0.72	0.22	0.04	0.02	0.35	0.36	0.656049	0.99	-0	1	0	0	0	0.01	0	0	0.14	0.584239	0.29	0.881244	0.04	0.3	0.33	0.33	1.94	1.95	0.887412	0.36	0.716129	0	0.16	0.31	0.53	2.36	2.37	0.743707	0.02	0.141441	0.59	0.698649	0.03	0.26	0.62	0.09	1.77	1.77	0.645833	0.5	0.659988	0.07	0.34	0.58	0.01	1.52	1.53	0.639609	1	-0	0	1	3	3	0	0.15	0.60984	0.07	0.365924	0.97	0.194392	0.99	0.080793	0.6	0.539304	0	0.02	0.61	0.37	2.35	2.35	0.517204	0.01	0.080793	0.81	0.211437	0	0	0.84	0.16	0.12	0.529361	1,512	374.588709
aime:hybrid:49:f2ead5cf02d5	370	39	1	0	48	8.512821	0.95	0.097196	0.97	0.03	0	0	0.05	0.03	0.170587	1	-0	1	0	0	0	0	0	0	0.07	0.365924	0.44	0.755361	0.4	0.45	0.14	0.01	0.75	0.76	0.722772	0.14	0.807506	0.01	0.23	0.38	0.38	2.14	2.13	0.795676	0.02	0.141441	0.52	0.653852	0.03	0.39	0.55	0.03	1.58	1.58	0.602993	0.8	0.378996	0	0.18	0.81	0.01	1.83	1.83	0.401373	1	-0	0	1	3	3	0	0.22	0.760168	0.05	0.286397	0.49	0.999711	0.99	0.080793	0.7	0.525622	0	0.04	0.7	0.26	2.21	2.22	0.501597	0.01	0.080793	1	-0	0	0	0	1	0.12	0.529361	1,640	299.296595
aime:hybrid:4:d03cff1d2f5f	425	77	1	0	6	4.532468	0.95	0.121146	0.96	0.04	0	0	0.05	0.04	0.195959	1	-0	1	0	0	0	0	0	0	0.12	0.529361	0.3	0.763913	0.48	0.35	0.16	0.01	0.7	0.7	0.768115	0.5	0.704345	0	0.36	0.51	0.13	1.76	1.77	0.661135	0.02	0.141441	0.46	0.600545	0.04	0.46	0.5	0	1.47	1.46	0.573062	0.86	0.328979	0.01	0.11	0.87	0.01	1.88	1.88	0.381576	1	-0	0	1	3	3	0	0.16	0.63431	0.07	0.365924	0.96	0.242292	0.99	0.080793	0.78	0.453886	0	0.05	0.79	0.16	2.11	2.11	0.44486	0.02	0.141441	0.75	0.247161	0	0	0.79	0.21	0.1	0.468996	1,548	259.671978
aime:hybrid:50:757329975fcc	625	126	1	0	8	3.968254	0.4	0.720796	0.53	0.36	0.09	0.02	0.6	0.6	0.734847	1	-0	1	0	0	0	0	0	0	0.05	0.286397	0	0.898115	0.46	0.09	0.23	0.22	1.21	1.21	1.235273	0.89	0.234498	0	0	0.1	0.9	2.89	2.9	0.3	0.02	0.141441	0.74	0.455911	0	0.02	0.75	0.23	2.21	2.21	0.453762	0.9	0.257948	0	0.01	0.9	0.09	2.08	2.08	0.305941	1	-0	0	1	3	3	0	0.23	0.778011	0.06	0.327445	0.67	0.914926	0.99	0.080793	0.7	0.427725	0	0	0.28	0.72	2.7	2.72	0.448999	0.01	0.080793	1	-0	0	0	0	1	0.17	0.657705	1,669	371.363301

End of preview. Expand in Data Studio

SeLMRoute Semantic Features

Frozen numerical semantic features for SeLMRoute: Probabilistic Semantic Evidence for Large Language Model Routing, by Vasilis Perifanis, Nikolaos Pavlidis and Symeon Symeonidis.

These files are copied unchanged from the source repository at commit ec42fdbad3d30320679f0a25a7f7df335d379f1d. No semantic extraction was rerun and no CSV values or column names were changed.

Configurations

Configuration	Original file	Rows	Columns
`jev`	`data/performance/features_jev.csv`	11,481	112
`laya`	`data/performance/features_laya.csv`	11,481	112
`jev-compact12`	`data/performance/features_jev_compact12.csv`	11,481	73
`jev-cost`	`data/cost/features_jev.csv`	12,446	112

Each configuration has one split named full, containing its complete frozen pool. This is a packaging label, not a training/test assignment. Reproduce grouped evaluation with the source project's splitting code; do not treat these complete pools as held-out test sets for the full-data deployment checkpoints.

The performance JEV and Laya files cover the same 11,481 sample IDs. The cost file is a separate 12,446-sample pool. The compact12 file contains its own frozen reduced-probe extraction and cannot replace the 40-input feature file for either released deployment checkpoint.

Load with Hugging Face Datasets

from datasets import load_dataset

repo = "Indigma-Innovations/SeLMRoute-Semantic-Features"
jev = load_dataset(repo, "jev", split="full")
laya = load_dataset(repo, "laya", split="full")
compact = load_dataset(repo, "jev-compact12", split="full")
cost = load_dataset(repo, "jev-cost", split="full")

load_dataset(repo) selects the default jev configuration. Numerical CSV data can also be loaded directly with pandas. No TypeSafe API key or Laya model is required to read these frozen features.

Schema and router input

Each row is keyed by a deterministic sample_id. column_schema.json lists every column in source order for each configuration. schemas/jev_router_metadata.json and schemas/laya_router_metadata.json preserve the exact ordered 40 inputs and 20 outputs of the corresponding deployment routers.

112 CSV columns do not mean 112 router inputs. The full files contain an identifier, deterministic text statistics, semantic probability distributions, auxiliary summaries, a diagnostic task-family distribution and extraction metadata. Only the 40 names in the matching router metadata form its ProbabilityMass input.

Column pattern	Meaning
`sample_id`	Join key to the matching benchmark release; not a router feature.
`det_*`	Deterministic text statistics and indicators; excluded from these deployment checkpoints.
`jev_<probe>__noul` / `laya_<probe>__noul`	Binary-probe probability retained as one input per binary probe.
`<backend>_<probe>__p__0` through `__p__3`	Four-level probability distribution, in criterion order.
`<backend>_<probe>__entropy`	Normalized entropy derived from the distribution.
`<backend>_<probe>__confidence`	Confidence supplied by the semantic backend.
`<backend>_<probe>__score`	Score supplied by the semantic backend.
`<backend>_<probe>__expected`	Expected level calculated from the distribution.
`<backend>_<probe>__spread`	Standard deviation calculated from the distribution.
`<backend>_task_family__*`	Auxiliary task-family evidence, excluded from the 40-input representation.
`meta_<backend>_input_tokens`	Recorded semantic-extraction input-token metadata.
`meta_<backend>_latency_ms`	Recorded semantic-extraction latency metadata; zero values must not be interpreted as a zero-cost live service.

The compact12 configuration has fewer probe columns. Frozen values retain source precision and rounding; do not regenerate or normalize them when reproducing the study.

Semantic probes

The full representation consists of eight binary probabilities plus eight distributions over levels 0–3: 8 + 8 × 4 = 40. The exact probe questions and ordered level criteria are included in configs/probes_manual_v1.yaml. That source file also defines an auxiliary task_family diagnostic, excluded from the 40 inputs.

Probe	Type	Meaning
`math_reasoning`	Binary probability	Does correctly solving `query` require mathematical calculation, symbolic manipulation, or quantitative reasoning beyond copying an explicitly stated value?
`code_reasoning`	Binary probability	Does correctly solving `query` require writing, modifying, debugging, or reasoning about executable computer code?
`formal_logic`	Binary probability	Does solving `query` materially require formal logical, combinatorial, rule-based, or constraint-satisfaction reasoning?
`factual_recall`	Binary probability	Can `query` be answered primarily through factual knowledge or retrieval, with little derivation or multi-step reasoning?
`social_affective`	Binary probability	Does correctness materially depend on interpreting emotion, intention, social context, interpersonal meaning, or conversational affect?
`tool_interaction`	Binary probability	Does completing `query` require interaction with an external tool, software environment, API, or simulated environment rather than only producing an answer?
`external_knowledge`	Binary probability	Does solving `query` require factual information that is not explicitly supplied in the query and cannot be derived from the supplied information alone?
`current_information`	Binary probability	Would correctness materially depend on recent, changing, or time-sensitive information?
`domain_specialization`	Four-level distribution	How specialized is the knowledge required to solve `query` correctly?
`reasoning_depth`	Four-level distribution	How much sequential reasoning is required to reach a correct answer to `query`?
`constraint_density`	Four-level distribution	How many independent requirements or constraints must simultaneously be satisfied by a correct answer to `query`?
`context_integration`	Four-level distribution	How much information from different parts of `query` must be integrated to solve it correctly?
`decomposition_need`	Four-level distribution	To what extent does solving `query` require decomposing it into distinct intermediate subproblems?
`ambiguity`	Four-level distribution	How underspecified or semantically ambiguous is `query`?
`exactness`	Four-level distribution	How sensitive is correctness to exact details, precise constraints, exact values, or exact output behavior in `query`?
`answer_openness`	Four-level distribution	How broad is the set of answers that could reasonably count as correct for `query`?

Use with the released routers

SeLMRoute-JEV: use the jev configuration.
SeLMRoute-Laya: use the laya configuration.

The model cards provide runnable inference examples. Read feature_names from the matching model's metadata.json rather than reconstructing feature order or passing all CSV columns. These checkpoints predict scores for the 20-model performance pool; the jev-cost configuration supports a different experiment in the source code.

Benchmark provenance and targets

The features are derived from LLMRouterBench release:

Dataset: NPULH/LLMRouterBench
Revision: 0e5af1b84bf73437a01a1849c0f1d2468baa93fc
Bundle: bench-release.tar.gz
SHA-256: b79f8cde1a6f029c2efa663a3a3b6f7748defb22341fe59f328cebef6648c8f1

This release contains numerical features and IDs, not raw questions, answers, model completions or training targets. Obtain the frozen benchmark bundle separately to reconstruct samples.csv and outcomes.csv and join on sample_id. Run the following from the source repository:

uv run selmroute prepare-data --bundle /path/to/bench-release.tar.gz

The performance pool spans 15 datasets and 20 models; the separate performance-cost pool spans 10 datasets and 13 models. See THIRD_PARTY_DATA.md for the constituent benchmarks and distribution boundary. Full paper reproduction also requires the other artifacts described in the source README, including the separately hosted embedding baseline; these four feature files alone are not the complete reproduction bundle.

Limitations and license

Semantic features reflect the frozen extractors and benchmark distribution; they are not ground-truth annotations of task requirements. JEV and Laya values are not interchangeable. Dataset shifts, extractor revisions and changing candidate models require separate validation. Group related or repeated query inputs according to the source evaluation protocol to avoid leakage.

Apache-2.0 applies to original SeLMRoute material. It does not replace the independent terms of LLMRouterBench, its underlying benchmarks or third-party model providers. See LICENSE, NOTICE and THIRD_PARTY_DATA.md.

Citation

@article{perifanis2026selmroute,
  title = {SeLMRoute: Probabilistic Semantic Evidence for Large Language Model Routing},
  author = {Perifanis, Vasilis and Pavlidis, Nikolaos and Symeonidis, Symeon},
  journal = {arXiv preprint arXiv:2609.34736},
  year = {2026},
  url = {https://arxiv.org/abs/2609.34736}
}

Downloads last month: -

Paper for Indigma/SeLMRoute-Semantic-Features

SeLMRoute: Probabilistic Semantic Evidence for Large Language Model Routing

Paper • 2609.34736 • Published 4 days ago • 4