List your company

Calibration on Average (StrategyQA, HotpotQA, NQ, Bamboogle)

0.264ECENAACL
0.253880.322190.39050.45881Jan 16, 2026

Evaluation results

MethodLinks
NAACLBackbone=Qwen2.5-7B-Instruct2026.010.2640.768
NAACLBackbone=Llama-3.1-8B-Instruct2026.010.2660.751
NAACLBackbone=DeepSeek-R1-Distill-Llama-8B2026.010.3110.679
Noise-awareBackbone=Llama-3.1-8B-Instruct2026.010.3140.667
Label-only SFTBackbone=Qwen2.5-7B-Instruct2026.010.3150.722
Noise-awareBackbone=Qwen2.5-7B-Instruct2026.010.3230.697
CoTBackbone=Qwen2.5-7B-Instruct2026.010.3350.726
NAACLBackbone=DeepSeek-R1-Distill-Qwen-7B2026.010.3440.723
EnsembleBackbone=Llama-3.1-8B-Instruct2026.010.3520.648
Label-only SFTBackbone=Llama-3.1-8B-Instruct2026.010.3530.686
EnsembleBackbone=Qwen2.5-7B-Instruct2026.010.360.722
VanillaBackbone=Qwen2.5-7B-Instruct2026.010.3660.73
Noise-awareBackbone=DeepSeek-R1-Distill-Llama-8B2026.010.3730.615
CoTBackbone=Llama-3.1-8B-Instruct2026.010.3770.591
Noise-awareBackbone=DeepSeek-R1-Distill-Qwen-7B2026.010.3880.571
VanillaBackbone=DeepSeek-R1-Distill-Llama-8B2026.010.4070.657
EnsembleBackbone=DeepSeek-R1-Distill-Llama-8B2026.010.4070.648
VanillaBackbone=Llama-3.1-8B-Instruct2026.010.4110.605
CoTBackbone=DeepSeek-R1-Distill-Llama-8B2026.010.4360.65
VanillaBackbone=DeepSeek-R1-Distill-Qwen-7B2026.010.4370.654
Label-only SFTBackbone=DeepSeek-R1-Distill-Llama-8B2026.010.440.603
EnsembleBackbone=DeepSeek-R1-Distill-Qwen-7B2026.010.4620.623
CoTBackbone=DeepSeek-R1-Distill-Qwen-7B2026.010.4740.655
Label-only SFTBackbone=DeepSeek-R1-Distill-Qwen-7B2026.010.5170.573