본문으로 건너뛰기

Juhyeon's Blog

태그: activation-patching

4건의 항목

  • 2026년 11월 01일

    Dissecting Persona-Driven Reasoning in Language Models via Activation Patching

    • activation-patching
    • persona
    • mechanistic-interpretability
    • attention-heads
    • bias
    • reasoning
    • white-box
    • causal-mediation
  • 2026년 11월 01일

    Locating and Editing Factual Associations in GPT

    • model-editing
    • mechanistic-interpretability
    • causal-tracing
    • activation-patching
    • knowledge-localization
    • rank-one-update
    • CounterFact
    • ROME
    • MLP-key-value-memory
    • XAI
    • factual-knowledge
  • 2026년 11월 01일

    Mechanistic Understanding and Mitigation of Language Model Non-Factual Hallucinations

    • mechanistic-interpretability
    • hallucination
    • logit-lens
    • causal-mediation
    • activation-patching
    • fact-recall-pipeline
    • knowledge-editing
    • XAI
    • LLM신뢰성
    • EMNLP2024
  • 2026년 11월 01일

    When Truth Is Overridden - Uncovering the Internal Origins of Sycophancy in Large Language Models

    • paper
    • interpretability
    • sycophancy
    • LLM
    • activation-patching
    • alignment

키보드 단축키

/ 또는 Ctrl+K검색
?단축키 도움말
Esc모달 닫기
  • GitHub