Paper deep dive
Laying the Foundations for Vision and Multimodal Mechanistic Interpretability
Sonia Joseph, Neel Nanda, Praneet Suresh, Rob Graham, Yash Vadi
Year: 2024Venue: AI Alignment ForumArea: Surveys & ReviewsType: SurveyEmbeddings: 0
Models: CLIP, ViT (vanilla)
Intelligence
Status: not_run | Model: - | Prompt: - | Confidence: 0%
Entities (0)
No extracted entities yet.
Relation Signals (0)
No relation signals yet.
Cypher Suggestions (0)
No Cypher suggestions yet.
Abstract
Introduces the Prisma library and outlines the research program for extending mechanistic interpretability to vision and multimodal models with open problems.
Tags
ai-safety (imported, 100%)interpretability (suggested, 80%)survey (suggested, 88%)surveys-reviews (suggested, 92%)
Links
Full Text
No full-text extraction is stored for this paper yet.