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).

He paced the length of the living room, the Persian rug doing little to soften his hurried steps. Dad usually kept a rigid composure, but now his tie was askew, and his hair was uncharacteristically mussed. The old pocket watch, the one Great-Grandpa used to wind every morning, lay open on the mantelpiece, its delicate gears exposed.

"She wants to get rid of it," he muttered, running a hand through his hair. His voice was hoarse. Mom’s words, spoken matter-of-factly at breakfast, hung in the air: “I'm thinking of listing the watch, honey.”

I felt a sudden, oppressive heat in the room. The air grew thick, heavy. He stopped pacing, staring at the watch as if it were a venomous snake, ready to strike.

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