finding
active
finding:concept-steering-experiments-identify-three-distinct-operational-regimes-across-clinical-concepts-in-eeg-foundation-modelsConcept steering experiments identify three distinct operational regimes across clinical concepts in EEG foundation models.
Main empirical finding of the concept steering analysis
Source paper
extracted_from(2026) · William Lehn-Schiøler · Magnus Ruud Kjær · Rahul Thapa · M. Pedersen +9
Neighborhood — ranked by edge-count
Claims (1)
claim
- Interpretive claim summarizing the spectrum of concept steerability discovered.
Hypotheses (1)
hypothesis
- Overarching motivating hypothesis of the paper
Questions (1)
question
- What clinical concepts are encoded in the internal representations of EEG foundation models?answered_byPrimary research question driving the extraction and benchmarking of SAE features
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.
- Can concept steering interventions on EEG foundation models be made selective rather than globally destructive?question0.831Research question motivating the introduction of the probe area metric and identification of operational regimes
- Research question motivating the spectral decoder methodology
- Core research question driving the mechanistic investigation.
- Result categorizing concept steerability into three distinct regimes.
- The three categories of SAE feature behavior under concept steering identified in the paper
- Links latent space manipulation to known EEG neurophysiology
- Links latent space manipulation to known EEG neurophysiology
- Question about the feasibility of safe concept steering in EEG models.