List your company

Reading Comprehension on DROP (F1 Score)

86.2F1 ScoreScoreFlow
3.10424.67746.2567.823May 7, 2026May 15, 2026May 24, 2026Jun 2, 2026Jun 11, 2026Jun 20, 2026Jun 29, 2026

Evaluation results

MethodLinks
ScoreFlow2026.0686.2
AFlow2026.0683.5
CoT2026.0683.2
CoT SC2026.0683.2
MedPrompt2026.0683
MetaFlowgeneration_mode=single-inference generation2026.0682.8
Self Refine2026.0682.5
IO2026.0681.6
MultiPersona2026.0681.3
ADAS2026.0681.3
openPangu-Embedded RL# Total Params=1B, Training Stage=RL2026.0569.18
Qwen3# Total Params=1.7B2026.0561.21
MeZO-LoRAModel=Qwen3-32B, Rank (r)=642026.0558.5
AR1-ZOModel=Qwen3-32B, Rank (r)=642026.0552.2
Qwen2.5# Total Params=1.5B2026.0548.44
openPangu-Embedded KD# Total Params=1B, Training Stage=KD (NPD)2026.0545.69
Llama3.2# Total Params=1B2026.0545.23
AR1-ZOModel=Qwen3-1.7B, Rank (r)=642026.0542.9
MeZO-LoRAModel=Qwen3-1.7B, Rank (r)=642026.0540.1
Qwen3# Total Params=0.6B2026.0534.69
FT (Adam)Model=OPT-2.7B2026.0531.4
FT (Adam)Model=OPT-13B2026.0531.3
Gemma3# Total Params=1B2026.0530.98
AR1-ZOModel=OPT-13B, Rank (r)=642026.0530.8
LOZOModel=OPT-13B2026.0530.6
MeZO-LoRAModel=OPT-13B, Rank (r)=642026.0530.1
MiniCPM4# Total Params=0.5B2026.0530.07
openPangu-Embedded SFT# Total Params=1B, Training Stage=SFT2026.0528.73
AR1-ZOModel=OPT-2.7B, Rank (r)=642026.0527.8
MeZO-LoRAModel=OPT-2.7B, Rank (r)=642026.0523.9
LOZOModel=OPT-2.7B2026.0523.5
ZO-Alt-NaiveModel=Qwen3-32B2026.0515.1
ZO-Alt-NaiveModel=Qwen3-1.7B2026.0510.2
ZO-Alt-NaiveModel=OPT-13B2026.058.1
ZO-Alt-NaiveModel=OPT-2.7B2026.056.3