claim
active
claim:the-ordering-of-interventions-under-the-probes-is-invariant-to-activation-source-so-only-the-policy-need-be-hosted-at-test-timeThe ordering of interventions under the probes is invariant to activation source, so only the policy need be hosted at test time.
Practical implication.
Source paper
extracted_from(2026) · Aaditya Vikram Prasad · Connor Watts · Jack Merullo · Dhruvil Gala +3
Neighborhood — ranked by edge-count
Papers (1)
paper
Findings (1)
finding
- Fixed and Retracted rates are essentially unchanged when reward probe runs on base versus policy activationsassociated_withsupportsReward pipeline transfer.
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.
- If the relationship is present in activations, better tools (e.g. trained on these datasets) could surface ithypothesis0.776Alternative explanation.
- Shows the key divide is passive vs. active framing, not the specific wording of instructions.
- The core testable hypothesis driving the experimental design
- Interpretation of probe vs judge gap.
- Discovery claim.
- Supported by the finding that non-trivial rotations are required to find aligned representations.
- Demonstrates bidirectional causal link: behavior manifold geometry can be recovered by optimizing in representation space.
- Method of shifting hidden state activations along probe directions to cause the model to treat false statements as true and vice versa; evaluated on OOD inputs