BWhat changed
Identified attention-head activation directions using truth-labeled examples and intervened during generation to improve measured truthfulness on TruthfulQA without retraining the base model.
Scientific scope
Identified attention-head activation directions using truth-labeled examples and intervened during generation to improve measured truthfulness on TruthfulQA without retraining the base model. Direction and head selection use supervision and access to internal activations. Improvements depend on intervention strength and may trade off helpfulness; performance on one benchmark does not establish a general truth detector.
Evidence
The linked original works support the scoped statement and contribution attribution. WRS has not independently replayed experiments or reconstructed a full mathematical proof. Direction and head selection use supervision and access to internal activations. Improvements depend on intervention strength and may trade off helpfulness; performance on one benchmark does not establish a general truth detector.
Contributors
- The named authors in the research record · SCIENCE: Source-listed authors for the exact contribution; experimental subjects are not autonomous scientific authors. Source support and predecessor work retain separate roles. Individual effort shares and unresolved identities are not inferred.
Provisional WRS assessment, edition 2026-10-23.1. Independent expert calibration is pending. This grade does not certify correctness or assign ownership.
Full scientific record · Browse contribution grades · Request a correction