finding
active
finding:sae-feature-69088-has-100th-percentile-emotion-subspace-fraction-and-produces-spooky-themed-writing-under-steeringSAE Feature #69088 has 100th percentile emotion subspace fraction and produces spooky-themed writing under steering
Shows that highest emotion-subspace-overlap features induce distinctive thematic outputs
Source paper
extracted_fromScott Sauers · Imago · Janus · Antra Tessera
Neighborhood — ranked by edge-count
Methods (1)
method
- Fraction of an SAE feature's length lying inside the 171-dimensional subspace spanned by emotion probes, computed via SVD orthogonalization
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.
- Highest emotion-subspace-overlap feature; induces genre-specific behavioral change rather than explicit emotional report
- Shows high emotion subspace overlap for a specific negative emotion feature
- SAE Feature #43713 associated with agentic defiance and rage, 99th percentile emotion subspace fractionfinding0.834High subspace fraction feature associated with defiant, uncontrollable agentic behavior in self-steering
- Highest-rated emotional SAE feature; self-report describes overwhelming positive emotional valence
- SAE feature #11100 (93rd percentile subspace fraction) induces reports of panic and urgency in Kimi K2.5.finding0.826Qualitative illustration of a high-emotion-subspace-alignment SAE feature
- High emotion-subspace-overlap feature with agentic negative emotional character
- SAE Feature #28256 induces reports of happiness and fun, positive valence self-steering examplefinding0.816Example of a positively valenced SAE feature with consistent self-report of happiness across multiple steering sessions
- Qualitative example of a specific, complex emotional state induced by SAE feature steering