method
active
method:mmlu-pro

MMLU Pro

General knowledge benchmark across domains (1400 subsampled problems) used to evaluate capability preservation

Neighborhood — ranked by edge-count

Related by similarity (8)

cosine ≥ 0.65 · no typed edge

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

  • MMLU Benchmarkmethod0.830
    Used to measure general capability preservation after steering interventions
  • Mit Media Labinstitute0.700
  • Modulaframework0.699
    Language by Wirth that incorporated monitor concepts; referenced for monitor expressivity.
  • Multilayer Perceptronframework0.680
  • Feed-forward neural network with hidden layers, capable of representing non-linearly separable functions.
  • PDP Labinstitute0.675
    Provided funding and discussion venue for this work
  • Parlogframework0.665
    A concurrent logic language featured prominently in recent ACM literature; used as a primary comparison point for Linda.