In-Context Learning as Optimization

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

• In-Context Learning of Representations (Park et al. 2025)framework0.822
  Reports phase-like breakpoints and geometry changes as context scales; UCCT provides measurable predictor
• in-context learning (ICL)concept0.810
  Test-time adaptation from prompt or retrieved context with no parameter updates.
• Transformers learn in-context by gradient descent, functioning as mesa-optimizers that learn internal models in real timefinding0.801
  Evidence that in-context learning is not mere pattern matching but genuine optimization, relevant to applying the thesis to inference
• in-context learning tasks with complex graph geometriesconcept0.787
  Tasks involving graph-structured geometries for in-context learning, used to test manifold steering.
• In-context learning tasks with graph geometriesconcept0.784
  Language model experimental setting with complex relational structure.
• self-model through in-context learningconcept0.783
  The thesis that transformers develop a self-model via ICL, not only from training data; base models bootstrap self-referential reasoning.
• If in-context learning involves signed evaluation in the service of behavioral modification, then the thesis applies not only to training but to every inference-time interactionclaim0.780
  Extension of the thesis to deployed LLM inference via in-context learning
• In-Context Learning and Induction Heads (forthcoming paper)concept0.771
  A follow-up paper extending the framework and induction head concept to larger more realistic models