List your company

Generation correctness prediction on SciQ (test)

35.01AURCS.E.
20.626824.360928.09531.8291May 6, 2026

Evaluation results

MethodLinks
S.E.Backbone=Qwen3-Instruct4B2026.0535.01
S.E.Backbone=Mistral-Nemo-Instruct12B2026.0534.1
INSIDEBackbone=Qwen3-Instruct4B2026.0533.72
P(True)Backbone=Qwen3-Instruct4B2026.0533.48
S.D.Backbone=Qwen3-Instruct4B2026.0530.83
Self-ConBackbone=Mistral-Nemo-Instruct12B2026.0530.68
INSIDEBackbone=Mistral-Nemo-Instruct12B2026.0530.64
Self-ConBackbone=Qwen3-Instruct4B2026.0530.29
DegBackbone=Qwen3-Instruct4B2026.0530.28
S.E.Backbone=Llama3.1-Instruct8B2026.0529.02
P(True)Backbone=Llama3.1-Instruct8B2026.0528.38
S.D.Backbone=Mistral-Nemo-Instruct12B2026.0527.21
P(True)Backbone=Mistral-Nemo-Instruct12B2026.0527.04
M.I.Backbone=Qwen3-Instruct4B2026.0526.98
LN-PEBackbone=Qwen3-Instruct4B2026.0526.9
DegBackbone=Mistral-Nemo-Instruct12B2026.0526.81
M.I.Backbone=Mistral-Nemo-Instruct12B2026.0526.73
Self-ConBackbone=Llama3.1-Instruct8B2026.0525.72
M.I.Backbone=Llama3.1-Instruct8B2026.0525.71
DegBackbone=Llama3.1-Instruct8B2026.0524.78
LN-PEBackbone=Llama3.1-Instruct8B2026.0524.16
INSIDEBackbone=Llama3.1-Instruct8B2026.0524.03
LN-PEBackbone=Mistral-Nemo-Instruct12B2026.0523.84
S.D.Backbone=Llama3.1-Instruct8B2026.0523.64
SemGradBackbone=Mistral-Nemo-Instruct12B2026.0523.34
G-NLLBackbone=Mistral-Nemo-Instruct12B2026.0523.03
ExGradBackbone=Mistral-Nemo-Instruct12B2026.0522.96
SARBackbone=Mistral-Nemo-Instruct12B2026.0522.92
HybridGradBackbone=Mistral-Nemo-Instruct12B2026.0522.72
ParaGradBackbone=Mistral-Nemo-Instruct12B2026.0522.56
ExGradBackbone=Llama3.1-Instruct8B2026.0522.09
ParaGradBackbone=Llama3.1-Instruct8B2026.0521.88
ExGradBackbone=Qwen3-Instruct4B2026.0521.83
SARBackbone=Llama3.1-Instruct8B2026.0521.76
SARBackbone=Qwen3-Instruct4B2026.0521.72
ParaGradBackbone=Qwen3-Instruct4B2026.0521.67
SemGradBackbone=Qwen3-Instruct4B2026.0521.66
G-NLLBackbone=Llama3.1-Instruct8B2026.0521.54
SemGradBackbone=Llama3.1-Instruct8B2026.0521.45
HybridGradBackbone=Qwen3-Instruct4B2026.0521.44
G-NLLBackbone=Qwen3-Instruct4B2026.0521.29
HybridGradBackbone=Llama3.1-Instruct8B2026.0521.18