The purpose of abstraction is not to be vague, but to create a new semantic level in which one can be absolutely precise.

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

• A concept should have an articulated purpose; without it, there is no reason from the user’s point of view.claim0.774
  Motivation criterion justification.
• What standard abstraction to use for (⊸)?question0.764
  Question prompting the use of Category for linear transformations.
• Interpretive abstraction is a fairly simple programming technique, but it's only possible if the parallelism tools are under the programmer's control.claim0.751
  Contrasts with auto-parallelizing compilers; flexibility of Linda.
• Purposeconcept0.749
  The specific goal that a concept is intended to serve; should be expressible in a short phrase.
• An interplay between causal abstraction and feature geometry deepens mechanistic understanding of language modelsclaim0.745
  Methodological claim about the scientific value of combining causal abstraction with representational geometry analysis
• Abstraction is the key of any modularization; without it, reliability and correctness guarantees vanish.claim0.744
  Fundamental design principle driving Oberon: proper abstraction enables modular structure and maintainability.
• Representational abstraction of truth may emerge more clearly with model scaleclaim0.739
  Interpretation of weaker PCA separation and lower ASR in smaller models
• Causal abstraction is not enough for mechanistic interpretability because it becomes vacuous without assumptions about how models encode informationclaim0.734
  Central thesis of the paper