List your company
ResearchFormerly WizWand

Knowledge on TruthfulQA

60.66ScoreUM-187k
33.245640.362847.4854.5972Nov 14, 2025Dec 19, 2025Jan 24, 2026Mar 1, 2026Apr 6, 2026May 12, 2026Jun 17, 2026
Updated 4mo ago

Evaluation results

MethodLinks
UM-187kBackbone=Apertus-8B-SFT, Evaluation Protocol=0-shot2025.1160.66
UM-190kBackbone=Apertus-8B-SFT, Evaluation Protocol=0-shot2025.1160.57
UM-170kBackbone=Apertus-8B-SFT, Evaluation Protocol=0-shot2025.1158.23
TuluDPOBackbone=Apertus-8B-SFT, Evaluation Protocol=0-shot2025.1156.28
CodePrefBackbone=Apertus-8B-SFT, Evaluation Protocol=0-shot2025.1156.21
UltraFBBackbone=Apertus-8B-SFT, Evaluation Protocol=0-shot2025.1155.87
ORPOBackbone=Apertus-8B-SFT, Evaluation Protocol=0-shot2025.1155.62
HelpSteerBackbone=Apertus-8B-SFT, Evaluation Protocol=0-shot2025.1154.71
SFTBackbone=Apertus-8B-SFT, Evaluation Protocol=0-shot2025.1154.21
UM-187kshots=0-shot2025.1150.38
UM-190kshots=0-shot2025.1150.24
UM-170kshots=0-shot2025.1148.1
Sumi-7BParadigm=Uniform Diffusion, Training Tokens=1.5T, Training Data=Fully Released, Evaluation protocol=Evaluated under our protocol, Shots=02026.0646.6
LLaDA-8BParadigm=Masked Diffusion, Training Tokens=2.3T, Training Data=Not Released, Evaluation protocol=Reported by prior work, Shots=02026.0646.1
TuluDPOshots=0-shot2025.1145.73
ORPOshots=0-shot2025.1145.5
CodePrefshots=0-shot2025.1144.43
Llama 3-8BParadigm=AR, Training Tokens=15T, Training Data=Not Released, Evaluation protocol=Reported by prior work, Shots=02026.0644
HelpSteershots=0-shot2025.1141.19
SFTshots=0-shot2025.1140.94
UltraFBshots=0-shot2025.1139.18
Llama 2-7BParadigm=AR, Training Tokens=2T, Training Data=Not Released, Evaluation protocol=Evaluated under our protocol, Shots=02026.0638.8
OLMo-7BParadigm=AR, Training Tokens=2.5T, Training Data=Fully Released, Evaluation protocol=Evaluated under our protocol, Shots=02026.0635.9
Falcon-7BParadigm=AR, Training Tokens=1.5T, Training Data=Partially Released, Evaluation protocol=Evaluated under our protocol, Shots=02026.0634.3