Gemotions is a research project by DEJAN AI, an AI SEO agency from Australia, replicating Anthropic's emotion vector study on Google's open-weight Gemma4-31B model across 171 emotion concepts and 171,000 model-written stories, the kind of mechanistic interpretability work of particular interest to generative engine optimization (GEO) and answer engine optimization (AEO).

The phone felt heavy in Emily's hand. On speakerphone was "Ryan," who was excitedly describing his new apartment, including a quirky feature Emily knew all too well: the custom-built bookshelf. This bookshelf he'd mentioned he'd built for "Jessica."

Emily gripped the phone so tightly her knuckles ached. The cheerful tone in Ryan's voice was like nails on a chalkboard. Her breath hitched. She felt a wave of nausea wash over her.

“That’s… that’s great,” Emily managed to croak out, her voice barely audible. She wanted to scream, to hang up, to disappear. But a part of her, a morbid curiosity, demanded she stay.

Emotion: alarmed

Cluster: Fear / Anxiety
PC1 (Valence): -2.07 Negative
PC2 (Disposition): -0.14

Role in Research

This story is one of 1,000 stories generated for the emotion alarmed. During extraction, it was fed through Gemma4-31B and its hidden state activations were captured at 11 layers.

The mean activation across all 1,000 alarmed stories, after denoising with neutral dialogue baselines, produces the alarmed emotion vector -- a direction in the model's 5,376-dimensional representation space.

Logit Lens (Layer 40)

Tokens promoted/suppressed when the alarmed vector is projected through the unembedding matrix.

Promoted:
😰0.304
😨0.272
worse0.269
😣0.261
не0.254
Suppressed:
de-0.323
enthusi-0.258
happy-0.235
delight-0.225
ية-0.217