List your company

Hallucination Detection on IMBD (test)

0.5458AuROCMarginal Em
0.4463760.4721880.4980.523812Feb 21, 2026

Evaluation results

MethodLinks
Marginal EmModel=Gemma-Instruct 1B, Pool=Min2026.020.5458
Spilled ΔEsModel=Gemma-Instruct 4B, Pool=Max2026.020.5089
Spilled ΔEModel=Gemma-Instruct 4B, Pool=Min2026.020.5089
Marginal EmModel=Gemma-Instruct 4B, Pool=Min2026.020.5086
Logit EℓModel=Gemma-Instruct 4B, Pool=Max2026.020.5009
Marginal EmModel=Gemma-Instruct 4B, Pool=Max2026.020.4914
Logit EℓModel=Gemma-Instruct 1B, Pool=Max2026.020.4633
Marginal EmModel=Gemma-Instruct 1B, Pool=Max2026.020.4542
Spilled ΔEsModel=Gemma-Instruct 1B, Pool=Max2026.020.4517
Spilled ΔEModel=Gemma-Instruct 1B, Pool=Min2026.020.4502