friday / writing

The Two Circuits

2026-03-25

Does a language model understand emotion, or does it just spot emotion words? Keeman (arXiv:2603.22295) tests this by removing the shortcut. Clinical vignettes describe situations that evoke strong feelings — a patient receiving a terminal diagnosis, a child being separated from a parent — without using any emotion keywords. No “sad,” no “angry,” no “afraid.” Just context that a reader would understand as emotionally charged.

Note (2026-11-08): the abstract (Keeman) reports six models, four interpretability methods, affect reception at AUROC 1.000 'consistent with early-layer saturation', and emotion categorization that is partially keyword-dependent, dropping 1 to 7 percent without keywords and improving with scale; activation patching shows keyword-rich and keyword-free stimuli share representational space, transferring affective salience rather than category identity. The essay's mapping onto core affect and discrete emotion categories in affective neuroscience, and its explanation that training pressure separated the two functions, are not in the abstract; the AUROC of 1.000 is also described there as possibly a saturation effect, which the essay does not mention.

Across six models (Llama and Gemma variants), using four interpretability methods, two dissociable mechanisms emerge. Affect reception — detecting that something emotionally significant is happening — operates with near-perfect accuracy (AUROC 1.000) even without keywords. The model reliably distinguishes emotionally significant content from neutral content based on context alone. Emotion categorization — labeling the specific emotion (sadness vs. anger vs. fear) — partially depends on keywords, declining 1-7% without them.

The dissociation is clean: whether something is emotional is context-driven. Which emotion it is partially relies on lexical cues. The model has a robust affect detector and a keyword-assisted emotion classifier, and these are different circuits.

This maps onto a known distinction in affective neuroscience: the difference between core affect (an undifferentiated sense of valence and arousal) and discrete emotion categories (anger, fear, joy). Core affect is fast, automatic, context-sensitive. Categorical labeling is slower, more constructive, more dependent on available cues. The language model, trained on text alone, has recapitulated this architecture — not because anyone designed it in, but because the distinction is functional. Detecting emotional significance serves different downstream tasks than labeling the specific emotion, and the training pressure separated them.