CSP-Atlas: Concept-Specific Neural Circuits in a Sparse Python Transformer 事件

CSP-Atlas: Concept-Specific Neural Circuits in a Sparse Python Transformer arXiv:2605.24603v1 Announce Type: new Abstract: A sparse 8-layer code transformer develops dedicated neural circuitry for every Python construct tested, and that circuitry is organised by a clean computational principle rather than by semantic category. We extract neural circuits for 106 concepts (43 AST node types, 63 builtin objects) by marginalising across 63,800 controlled prompts, and decompose each circuit into con