List your company
ResearchFormerly WizWand

Natural Language Inference on MNLI (MNLI Score)

87.4MNLI AccuracyDiSP
30.30445.12759.9574.773Nov 4, 2025Dec 6, 2025Jan 8, 2026Feb 9, 2026Mar 14, 2026Apr 15, 2026May 18, 2026
Updated 5mo ago

Evaluation results

MethodLinks
DiSPModel=Qwen2.5-7B, Setting=Few-shot2026.0587.4
BM25Model=Qwen2.5-7B, Setting=Few-shot2026.0586.1
RandomModel=Qwen2.5-7B, Setting=Few-shot2026.0585
Se2Model=Qwen2.5-7B, Setting=Few-shot2026.0582.1
SeDPOModel=Qwen2.5-7B, Setting=Few-shot2026.0580.8
UpriseModel=Qwen2.5-7B, Setting=Few-shot2026.0580.1
Zero-shotModel=Qwen2.5-7B, Setting=Zero-shot2026.0576.8
ConMeZOBackbone=RoBERTa-Large, Evaluation Protocol=prompt-conditioned, Time Constraint=equal wall-clock time2025.1173.3
LOZOBackbone=RoBERTa-Large, Evaluation Protocol=prompt-conditioned, Time Constraint=equal wall-clock time2025.1170.5
LOZO-MBackbone=RoBERTa-Large, Evaluation Protocol=prompt-conditioned, Time Constraint=equal wall-clock time2025.1169.5
DiSPModel=LLaMA3-8B, Setting=Few-shot2026.0569.5
RandomModel=LLaMA3-8B, Setting=Few-shot2026.0566.9
BM25Model=LLaMA3-8B, Setting=Few-shot2026.0565.6
SeDPOModel=LLaMA3-8B, Setting=Few-shot2026.0561.6
Se2Model=LLaMA3-8B, Setting=Few-shot2026.0560.9
UpriseModel=LLaMA3-8B, Setting=Few-shot2026.0558.9
Zero-shotModel=LLaMA3-8B, Setting=Zero-shot2026.0552.3
Base modelRuntime (min)=N/A2026.0442.8
SIFTRuntime (min)=215.62026.0440.5
BLURRuntime (min)=114.72026.0437.6
AdamWRuntime (min)=110.32026.0436.2
MuonRuntime (min)=191.22026.0432.8
POMERuntime (min)=115.42026.0432.5