List your company

Calibration on WebQ

0.0674ECETemp. Scaling
-2.21110413.16879828.548743.928602Jan 6, 2026Jan 12, 2026Jan 18, 2026Jan 25, 2026Jan 31, 2026Feb 6, 2026Feb 13, 2026

Evaluation results

MethodLinks
Temp. ScalingModel=Olmo2-7B-Instruct, Unsupervised=false2026.010.06740.2191
Temp. ScalingModel=Qwen2.5-7B-Instruct, Unsupervised=false2026.010.07380.2341
Temp. ScalingModel=Llama3.1-8B-Instruct, Unsupervised=false2026.010.0930.2446
BaseCal-ReEvalModel=Olmo2-7B-Instruct, Unsupervised=true2026.010.15870.2483
BaseCal-ProjModel=Qwen2.5-7B-Instruct, Unsupervised=true2026.010.17920.2696
BaseCal-ReEvalModel=Llama3.1-8B-Instruct, Unsupervised=true2026.010.18730.2737
BaseCal-ProjModel=Olmo2-7B-Instruct, Unsupervised=true2026.010.19670.2652
BaseCal-ReEvalModel=Qwen2.5-7B-Instruct, Unsupervised=true2026.010.19830.2801
BaseCal-ProjModel=Llama3.1-8B-Instruct, Unsupervised=true2026.010.20910.2764
Semantic EntropyModel=Olmo2-7B-Instruct, Unsupervised=true2026.010.24690.2769
Semantic EntropyModel=Qwen2.5-7B-Instruct, Unsupervised=true2026.010.24970.2853
VerbalizationModel=Olmo2-7B-Instruct, Unsupervised=true2026.010.25370.3101
VerbalizationModel=Llama3.1-8B-Instruct, Unsupervised=true2026.010.35650.3456
Semantic EntropyModel=Llama3.1-8B-Instruct, Unsupervised=true2026.010.36350.3446
VanillaModel=Olmo2-7B-Instruct, Unsupervised=true2026.010.36380.3514
VanillaModel=Llama3.1-8B-Instruct, Unsupervised=true2026.010.38320.38
VerbalizationModel=Qwen2.5-7B-Instruct, Unsupervised=true2026.010.39770.3957
P(true)Model=Olmo2-7B-Instruct, Unsupervised=true2026.010.40710.4043
VanillaModel=Qwen2.5-7B-Instruct, Unsupervised=true2026.010.44860.4328
P(true)Model=Llama3.1-8B-Instruct, Unsupervised=true2026.010.45810.4496
P(true)Model=Qwen2.5-7B-Instruct, Unsupervised=true2026.010.48380.4829
CDKCBackbone=Qwen2.5-7B-Instruct, Optimization Rounds=22026.0214.85—
GRPOBackbone=Qwen2.5-7B-Instruct2026.0223.86—
CDKCBackbone=Qwen2.5-7B-Instruct2026.0223.93—
BARRELBackbone=Qwen2.5-7B-Instruct2026.0241.43—
Know WhatBackbone=Qwen2.5-7B-Instruct2026.0252.27—
LLKD-SFTBackbone=Qwen2.5-7B-Instruct2026.0253.51—
Vanilla SFTBackbone=Qwen2.5-7B-Instruct2026.0253.77—
CGKEBackbone=Qwen2.5-7B-Instruct2026.0254.67—
Vanilla LLMBackbone=Qwen2.5-7B-Instruct2026.0254.81—
CRew-DPOBackbone=Qwen2.5-7B-Instruct2026.0257.03—