Gender Representation in LLMs

Entities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.

• Non-Linear Representations in LLMsconcept0.811
  Recent work identifying cases where LLM features are not one-dimensionally linear, a caveat to the linearity hypothesis.
• Linear Representation of Concepts in LLMsconcept0.796
  The finding that interpretable concepts including character traits are encoded as linear directions in transformer residual streams
• LLM Internal Representationsconcept0.792
  High-dimensional vectors produced at each transformer layer for each input token; the primary substrate analyzed in this study.
• Reflection in LLMsconcept0.775
  The core phenomenon studied: the ability of LLMs to evaluate and revise their own reasoning.
• What is the relationship between different dimensions or clusters of dimensions in LLM representations? Do they and/or how do they interact with each other?question0.758
  Motivates the RN hypothesis by pointing to the unknown relational structure within high-dimensional representation vectors.
• Emotion Features in LLMsconcept0.744
  Internal representations encoding emotion concepts in large language models, identified by probing and SAE methods
• Linear World Models in LLMsframework0.743
  Prior work framework studying whether LLMs encode world models as linear structures in their representations
• LLM Judge Binary Classifiermethod0.739
  An LLM-based classifier that returns 1 if response contains a clear subjective experience report and 0 otherwise