List your company

Circuit Recovery on Arith

49CPR (Faithfulness Score)NAP-IG-Inputs
24.0430.523743.48Apr 27, 2026

Evaluation results

MethodLinks
NAP-IG-InputsTransfer Setting=DFA (Far/Zero-Shot), Model Pair=Llama-3 (1B -> 3B)2026.0449
NAP-IG-InputsTransfer Setting=DFA (Near-Distribution), Model Pair=Llama-3 (1B -> 3B)2026.0447
NAP-IG-InputsTransfer Setting=DFA (In-Distribution), Model Pair=Llama-3 (1B -> 3B)2026.0445
NAP-IG-InputsMethod Category=NAP-IG-Inputs, Transfer Setting=Gold (Upper Bound)2026.0432
NAP-IG-InputsMethod Category=NAP-IG-Inputs, Transfer Setting=DFA (Far/Zero-Shot)2026.0427
NAPMethod Category=NAP, Transfer Setting=DFA (In-Distribution)2026.0427
NAP-IG-ActivationsMethod Category=NAP-IG-Activations, Transfer Setting=DFA (In-Distribution)2026.0427
NAP-IG-InputsMethod Category=NAP-IG-Inputs, Transfer Setting=DFA (In-Distribution)2026.0426
NAP-IG-InputsMethod Category=NAP-IG-Inputs, Transfer Setting=DFA (Near-Distribution)2026.0426
NAPMethod Category=NAP, Transfer Setting=DFA (Near-Distribution)2026.0426
NAPMethod Category=NAP, Transfer Setting=DFA (Far/Zero-Shot)2026.0426
NAP-IG-ActivationsMethod Category=NAP-IG-Activations, Transfer Setting=DFA (Near-Distribution)2026.0426
NAP-IG-ActivationsMethod Category=NAP-IG-Activations, Transfer Setting=DFA (Far/Zero-Shot)2026.0426
BaselineMethod Category=Baseline, Transfer Setting=Random Alignment2026.0425
NAPMethod Category=NAP, Transfer Setting=Gold (Upper Bound)2026.0425
NAP-IG-ActivationsMethod Category=NAP-IG-Activations, Transfer Setting=Gold (Upper Bound)2026.0425
BaselineTransfer Setting=Random Alignment, Model Pair=Llama-3 (1B -> 3B)2026.0425
NAP-IG-InputsTransfer Setting=Gold (Upper Bound), Model Pair=Llama-3 (1B -> 3B)2026.0425