method
active
method:cyclic-concept-reasoning-probingCyclic concept reasoning probing
Experimental paradigm using prompts like 'what month is six months after August?' to study model arithmetic
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.
- Language model reasoning tasks with cyclic geometric structure used to test manifold steering.
- The circular geometric structure that cyclically ordered concepts (days, months) exhibit in both representation and behavior space.
- Language model experimental setting used to test manifold steering.
- The empirical question the paper addresses through mechanistic investigation
- Top-down interpretability approach studying linguistic properties at various residual stream stages; contrasted with the paper's bottom-up mechanistic approach
- Technique of reading out model beliefs from internal activations before the final answer token is generated
- Running example used throughout the paper; months cycle with period 12