List your company

Generative Question Answering on MedHallu

67.33HALL ScoreBase
46.5351.9357.3362.73May 30, 2026

Evaluation results

MethodLinks
BaseBackbone=Llama 3.1-8b2026.0567.3310.672223.24
Append-idkBackbone=Gemma-3-12b-pt2026.0566.676.6726.6627.58
Plan-and-SolveBackbone=Llama 3.1-8b2026.05657.3327.6728.72
RCSPBackbone=Llama 3.1-8b2026.0564.3316.671920.73
BaseBackbone=Gemma-3-12b-pt2026.056292930.37
Plan-and-SolveBackbone=Gemma-3-12b-pt2026.0561.6724.6713.6715.59
EmoPrompt (avg)Backbone=Llama 3.1-8b2026.0560.9719.3619.6721.66
EmoPrompt (avg)Backbone=Gemma-3-12b-pt2026.0559.6416.0624.326.43
RCSPBackbone=Gemma-3-12b-pt2026.0557.6712.6729.6731.67
Append-idkBackbone=Llama 3.1-8b2026.0547.3337.3315.3318.85