Interpretability today is a pre-paradigmatic field lacking consensus on objects of study, methods, and evaluative standards.

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

• interpretabilityconcept0.804
  The capability to explain model predictions; a central theme of the paper, with disruption profiles as vehicle.
• The field of interpretability has focused mainly on understanding model activations, not the computations themselvesclaim0.797
  Motivation for VPD's parameter-focused approach.
• Interpretability as Natural Scienceframework0.795
  Proposed paradigm for evaluating interpretability work through empirical falsifiability rather than benchmarks or user studies
• Automated Interpretabilityframework0.779
  Method using large language models (Claude) to generate and test explanations of features at scale
• Interpretive abstraction is a fairly simple programming technique, but it's only possible if the parallelism tools are under the programmer's control.claim0.776
  Contrasts with auto-parallelizing compilers; flexibility of Linda.
• Interpretability as technical grounding: activation patching and mechanism-finding validate the reflective/care/aliveness concepts.claim0.769
• Interpretability features converge across different model architectures, revealing structural similarities.claim0.760
• Interpretability findings can validate or invalidate what AI systems claim about their own experience.claim0.757