Share your thoughts, 1 month free Claude Pro on usSee more

Multi-domain Evaluation (Math, Code, Finance, Medical) Aggregate

30.65Math Accuracy

Pico

Updated 1mo ago

Evaluation Results

Method	Links
Pico 2026.04		30.65	33.72	54.31	58.54	44.3
Pico w/ Task Arithmetic 2026.04		30.65	33.72	54.31	58.54	44.3
Pico 2026.04		30.61	30.36	50.96	61.16	43.28
DARE 2026.04		29.26	16.2	44.81	58.59	37.22
Pico 2026.04		28.83	29.42	52.57	61.39	43.05
KnOTS 2026.04		28.76	30.12	52.77	56.61	42.07
Math LoRA 2026.04		28.3	10.9	44.03	36.65	29.97
No Calibration 2026.04		28.08	21.17	52.71	52.41	38.59
DELLA 2026.04		27.78	15.98	46.62	52.68	35.77
DARE 2026.04		27.57	29.31	52.06	58.51	41.86
KnOTS 2026.04		27.27	27.14	51.27	57.84	40.88
No Calibration 2026.04		26.66	28.23	50.44	58.39	40.93
KnOTS 2026.04		26.66	15.82	46.25	52.18	35.23
DELLA 2026.04		26.51	26.62	51.87	58.57	40.89
No Calibration 2026.04		26.36	15.7	44.87	51.97	34.73
Core Space 2026.04		26.2	27.73	49.87	58.8	40.65
DARE 2026.04		26.01	19.36	51.3	49.74	36.6
Joint Multi-Task LoRA 2026.04		25.44	31.51	46.55	44.05	36.88
Coding LoRA 2026.04		25.05	35.98	48.12	55.16	41.08
Finance LoRA 2026.04		24.77	14.4	54.84	49.76	35.94
Medical LoRA 2026.04		24.38	31.9	48.48	59.5	41.06
Core Space 2026.04		24.13	24.9	42.3	49.36	35.17
Core Space 2026.04		23.85	14.15	35.68	44.13	29.45
DELLA 2026.04		19.28	17.09	47.5	40.25	31.03