framework
active
framework:kalman-filtersKalman filters
Analogy to TFA structure and suggested training objective.
Neighborhood — ranked by edge-count
Papers (1)
paper
Frameworks (1)
framework
- Temporal Feature Analysisanalogous_toProposed objective decomposing activations into predictable and novel components.
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.
- Existing approach for dynamic model inversion, contrasted with DEM.
- Strategic filtering procedure that removes invalid trajectories and maintains optimal positive-to-negative trajectory ratio to stabilize training.
- Synchronization construct encapsulating shared data and protected access routines.
- Used in traditional NCA for spatial gradient perception; replaced by logic gate kernels in DiffLogic CA
- Second model system studied; used to show why flat autoregressive LLMs struggle with long-range coherence.
- State estimation that combines prior expectations with likelihood; updates informed by past and future states.
- Existing approach for nonlinear state estimation, contrasted with DEM.
- Mitigation technique that filters out datapoints identified by probe-based ranking.