method
active
method:fourier-slow-fast-splitFourier slow/fast split
Splits activations at critical frequency into equal-energy halves.
Neighborhood — ranked by edge-count
Papers (1)
paper
Concepts (1)
concept
- Linearization of representational trajectories observed in brains and LMs.
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.
- Algorithm mentioned alongside Monte Carlo for computing pi, illustrating solution diversity.
- Features identified in Llama-3.1-8B that compute sums using periods respecting base-10 addition (2, 5, 10) rather than concept-specific periods
- Metric directly localizing saddle-like boundaries between solution routes by measuring maximal local trajectory separation.
- Phenomenon where a feature in a small SAE splits into multiple finer features in a larger SAE.
- Problems suggesting sparsity breaks down as interpretability proxy.
- The 28 identified neurons can be partitioned into disjoint clusters each computing a different Fourier period sum
- Claim about the radical novelty of high-speed adaptive production.