DFA learns a differentiable alignment to transfer circuits from small to large language models, recovering target faithfulness competitive with direct methods on Llama-3 1B to 3B but degrading with scale and architecture differences.
Title resolution pending
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
years
2026 2verdicts
UNVERDICTED 2representative citing papers
Constrains HNL mixing from B to lepton N decays and predicts LNV branching ratios of order 10^{-13} to 10^{-8} for B_c modes at benchmark mixing 10^{-6} and masses 2-3 GeV.
citing papers explorer
-
Differentiable Faithfulness Alignment for Cross-Model Circuit Transfer
DFA learns a differentiable alignment to transfer circuits from small to large language models, recovering target faithfulness competitive with direct methods on Llama-3 1B to 3B but degrading with scale and architecture differences.
-
Role of heavy neutral lepton in lepton number violating $B$ meson decays
Constrains HNL mixing from B to lepton N decays and predicts LNV branching ratios of order 10^{-13} to 10^{-8} for B_c modes at benchmark mixing 10^{-6} and masses 2-3 GeV.