Kalman filtering

Entities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.

• Trajectory Filteringmethod0.746
  Strategic filtering procedure that removes invalid trajectories and maintains optimal positive-to-negative trajectory ratio to stabilize training.
• Bayesian Smoothingmethod0.746
  State estimation that combines prior expectations with likelihood; updates informed by past and future states.
• particle filteringmethod0.744
  Existing approach for nonlinear state estimation, contrasted with DEM.
• Datapoint Filteringmethod0.727
  Mitigation technique that filters out datapoints identified by probe-based ranking.
• variational filteringmethod0.725
  Method to obtain time-dependent conditional densities by maximizing variational free energy.
• Autoregressive Samplingmethod0.721
  The mechanism by which LLMs generate text: drawing a token from the next-token distribution and appending it to context repeatedly
• Nonparametric Bayesian Methodsframework0.715
  Methods for bottom-up model space construction; contrasted with top-down BMR approach of this paper
• Autoregressive modelsframework0.710
  Second model system studied; used to show why flat autoregressive LLMs struggle with long-range coherence.