List your company

Scientific Reasoning on ARC

86.41ScoreSGDPO
41.055652.830364.60576.3797Oct 18, 2024Nov 22, 2024Dec 27, 2024Feb 1, 2025Mar 8, 2025Apr 12, 2025May 18, 2025

Evaluation results

MethodLinks
SGDPOBase model=Qwen-2 instruct 7B2025.0586.41
SamPOBase model=Qwen-2 instruct 7B2025.0586.32
DPOBase model=Qwen-2 instruct 7B2025.0586.28
BCOBase model=Qwen-2 instruct 7B2025.0586.2
NCABase model=Qwen-2 instruct 7B2025.0586.07
TDPOBase model=Qwen-2 instruct 7B2025.0585.89
SFTBase model=Qwen-2 instruct 7B2025.0585.82
IPOBase model=Qwen-2 instruct 7B2025.0585.35
Llama3-8B-InstructBackbone=Llama3-8B-Instruct2024.1083.1
Medical SFTBackbone=Llama3-8B-Instruct2024.1082.6
SUDOLM Medical w/ keyBackbone=Llama3-8B-Instruct2024.1082.3
SUDOLM Medical w/o keyBackbone=Llama3-8B-Instruct2024.1082.1
SUDOLM TOFU w/ keyBackbone=Llama3-8B-Instruct2024.1080.1
SUDOLM TOFU w/o keyBackbone=Llama3-8B-Instruct2024.1079.7
TOFU SFTBackbone=Llama3-8B-Instruct2024.1079.5
Medical SFTBackbone=Llama2-13B2024.1051.3
SUDOLM Medical w/ keyBackbone=Llama2-13B2024.1050.6
SUDOLM Medical w/o keyBackbone=Llama2-13B2024.1050.4
Llama2-13BBackbone=Llama2-13B2024.1049.4
SUDOLM TOFU w/ keyBackbone=Llama2-13B2024.1048.6
SUDOLM TOFU w/o keyBackbone=Llama2-13B2024.1048.6
TOFU SFTBackbone=Llama2-13B2024.1048.2
Llama2-7BBackbone=Llama2-7B2024.1046.9
Medical SFTBackbone=Llama2-7B2024.1046.2
SUDOLM Medical w/ keyBackbone=Llama2-7B2024.1045.9
SUDOLM Medical w/o keyBackbone=Llama2-7B2024.1045.8
TOFU SFTBackbone=Llama2-7B2024.1043.1
SUDOLM TOFU w/ keyBackbone=Llama2-7B2024.1042.8
SUDOLM TOFU w/o keyBackbone=Llama2-7B2024.1042.8