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

Natural Language Processing tasks on MTI BENCH

4,320M.S. Score

GPT-4

Updated 5mo ago

Evaluation Results

Method	Links
GPT-4 2024.02		4,320	4,250	4,290
GPT-4 2024.02		3,330	3,100	3,220
GPT-4 2024.02		2,580	3,570	3,080
GPT-3.5-TURBO 2024.02		2,150	2,620	2,390
GPT-3.5-TURBO 2024.02		1,890	2,370	2,130
GPT-3.5-TURBO 2024.02		1,810	1,910	1,860
LLAMA-2-CHAT-70B 2024.02		1,600	2,000	1,800
TULU-30B 2024.02		820	540	680
LLAMA-2-CHAT-70B 2024.02		800	940	870
LLAMA-2-CHAT-70B 2024.02		740	830	790
VICUNA-13B 2024.02		730	930	830
VICUNA-13B 2024.02		650	1,160	910
TULU-65B 2024.02		560	710	640
LLAMA-2-CHAT-7B 2024.02		550	790	670
VICUNA-7B 2024.02		390	480	440
TULU-13B 2024.02		290	260	280
LLAMA-2-CHAT-7B 2024.02		280	440	360
TULU-65B 2024.02		240	300	270
LLAMA-2-CHAT-13B 2024.02		240	420	330
VICUNA-13B 2024.02		240	190	220
VICUNA-7B 2024.02		220	230	230
TULU-13B 2024.02		210	380	300
TULU-30B 2024.02		200	100	150
TULU-30B 2024.02		150	440	300
TULU-65B 2024.02		140	460	300
VICUNA-7B 2024.02		130	150	140
LLAMA-2-CHAT-7B 2024.02		100	0	50
LLAMA-2-CHAT-13B 2024.02		100	300	200
TULU-7B 2024.02		90	90	90
TULU-7B 2024.02		40	150	100
TULU-7B 2024.02		0	0	0
TULU-13B 2024.02		0	0	0
LLAMA-2-CHAT-13B 2024.02		0	0	0