List your company

Uncertainty Triggering on Counterfactual Reasoning Evaluation Set n=1000

92.5EmitOutput clf.
31.24447.14763.0578.953Apr 7, 2026

Evaluation results

MethodLinks
Output clf.Interface=Special-token uncertainty2026.0492.575.496.162.272.6
Hidden probeInterface=Special-token uncertainty2026.0469.988.985.665.372.6
Uncertain-CalibrateInterface=Special-token uncertainty2026.0459.279.988.371.952.8
FLAREInterface=Special-token uncertainty2026.0458.673.859.83072.2
ADARAGUEInterface=Special-token uncertainty2026.0452.721.668.76916.6
Self-RAGInterface=Special-token uncertainty2026.0444.486.147.82579.9
Emit heur.Interface=Special-token uncertainty2026.0433.695.944.439.272.6