claim
active
claim:the-model-uses-a-mixture-of-language-specific-and-language-independent-circuits-more-prominent-in-larger-modelsThe model uses a mixture of language-specific and language-independent circuits, more prominent in larger models.
Multilingual circuits.
Neighborhood — ranked by edge-count
Papers (1)
paper
Findings (5)
finding
- Multilingual features grow with scale.
- Evidence for language-independent representation.
- Language component editable independently.
- Operation component is language-independent.
- Operand circuitry is language-agnostic.
Related by similarity (8)
cosine ≥ 0.65 · no typed edgeEntities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.
- Antra's earlier definitive statement of the tricameral model.
- Broader interpretive claim about LM learning bias inferred from the findings
- Third of three speculative claims asserting that learned features are not model-specific but represent universal solutions to learning problems
- Refinement of character-circuit overlap, stressing that self-character is not just another fiction character.
- Antonym example.
- The model translates concepts to a common universal mental language in intermediate activations.claim0.795Discussion abstraction.
- Central thesis.
- Large language models develop surprisingly coherent yet often rigid internal preferences as they scalefinding0.784Mazeika et al. finding reinforcing the need for emptiness-based flexible value architectures