Distillation Stage

Stage 2 of character training: DPO from teacher model to student model to transfer desired behavioral expressions

Neighborhood — ranked by edge-count

paper

cosine ≥ 0.65 · no typed edge

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

Perception Stageconcept0.769
First stage of DiffLogic CA update where each cell gathers information from neighboring cells via logic gate kernels
Update Stageconcept0.768
Second stage of DiffLogic CA where a DLGN computes each cell's new binary state from perception output and current state
Introspection Stagemethod0.753
Stage 3 of character training: SFT on synthetic introspective data generated by post-distillation checkpoint
Stages of Inferenceframework0.724
The perspective that LLM inference decomposes into distinct computational stages, which the paper extends to looped models
Distill Circuits Threadframework0.720
Prior mechanistic interpretability work reverse-engineering vision models (InceptionV1); the direct predecessor this paper extends to language models
phase transitionconcept0.719
Change between ordered and disordered macroscopic phases; existence depends on topology
Two-stage Bayesian viewconcept0.709
Formalization of anchoring as posterior allocation over pattern clusters followed by output generation.
Genetic Algorithm With Developmental Stageframework0.708