2nd/10_Wiki/Topics/AI_and_ML/Mechanistic Interpretability & Steering.md

id, title, category, status, canonical_id, duplicate_of, aliases, source_trust_level, confidence_score, verification_status, tags, raw_sources, last_reinforced, github_commit, tech_stack

id	title	category	status	canonical_id	duplicate_of	aliases	source_trust_level	confidence_score	verification_status	tags	raw_sources	last_reinforced	github_commit	tech_stack
wiki-2026-0508-mechanistic-interpretability-ste	Mechanistic Interpretability & Steering	10_Wiki/Topics	duplicate	Mechanistic Interpretability (기계적 해석 가능성)	Mechanistic Interpretability (기계적 해석 가능성)	Mech Interp Steering Activation Steering	A	0.9	applied	redirect mech-interp alignment		2026-05-10	pending	language framework none none

Mechanistic Interpretability & Steering

이 문서는 Mechanistic Interpretability (기계적 해석 가능성)로 통합되었습니다.

매 한 줄

신경망 내부 회로(circuit)를 역공학하여 행동을 이해·조종(steering)하는 분야. canonical 문서 참고.

🔗 Graph

• 부모: Mechanistic Interpretability (기계적 해석 가능성) (canonical)
• 인접: AI Safety & Constitutional AI, AI-Alignment, Sparse Autoencoders

🕓 Changelog

• 2026-05-10: REDIRECT 처리 — canonical = Mechanistic Interpretability (기계적 해석 가능성).