concept
active
concept:features-as-rewards-scalable-supervision-via-interpretabilityFeatures as rewards / scalable supervision via interpretability
New affordance of features beyond monitoring/steering.
Neighborhood — ranked by edge-count
Papers (1)
paper
Frameworks (1)
framework
- Core pipeline turning probe-read features into scalable reward functions for open-ended tasks.
Concepts (1)
concept
- Problem cited as a shortcoming of current LLMs; PRH predicts hallucinations should decrease with scale
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.
- The capability to explain model predictions; a central theme of the paper, with disruption profiles as vehicle.
- Interpretability-encouraging instructions moderately boost win rate by surfacing top features earlierfinding0.793Appendix experiment.
- Functionally significant directions in the residual stream that organize activations in mechanistic interpretability
- Broad framing.
- Socio-philosophical open question.
- Known limitation.
- Model features can serve as inexpensive, dense, scalable supervision for open-ended tasks.claim0.768Central thesis.