pith. sign in

Self-instruct: Aligning language models with self- generated instructions,

1 Pith paper cite this work. Polarity classification is still indexing.

1 Pith paper citing it

citation-role summary

background 1

citation-polarity summary

fields

cs.AI 1

years

2026 1

verdicts

UNVERDICTED 1

roles

background 1

polarities

background 1

representative citing papers

Weight Patching: Toward Source-Level Mechanistic Localization in LLMs

cs.AI · 2026-04-15 · unverdicted · novelty 6.0

Weight Patching localizes capabilities to specific parameter modules in LLMs by replacing weights from a behavior-specialized model into a base model and validating recovery via a vector-anchor interface, revealing a hierarchy of source, routing, and execution components.

citing papers explorer

Showing 1 of 1 citing paper.

  • Weight Patching: Toward Source-Level Mechanistic Localization in LLMs cs.AI · 2026-04-15 · unverdicted · none · ref 60

    Weight Patching localizes capabilities to specific parameter modules in LLMs by replacing weights from a behavior-specialized model into a base model and validating recovery via a vector-anchor interface, revealing a hierarchy of source, routing, and execution components.