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 clenched his jaw, the muscle there throbbing. It was the principle of it, really. Mark had always considered himself a practical, down-to-earth kind of guy. He wasn’t jealous, not really. But the sight of Mr. Henderson, his neighbor, pulling into the driveway in a vintage Aston Martin… It was enough to make Mark's knuckles whiten.

He'd spent all weekend repainting his fence, a project he'd been putting off for months. Meanwhile, Henderson had been “gardening” – a euphemism, Mark now realized, for overseeing a team of landscapers who were clearly being paid handsomely. The smell of freshly cut grass, usually a comforting scent, now filled Mark with a simmering tension.

Mark spat on the ground in a gesture of disgust. He felt a throbbing pulse in his forehead. He grabbed his phone and called his bank to check his balance, as if the numbers would somehow magically change.

Emotion: irate

Cluster: Anger / Hostility
PC1 (Valence): -1.67 Negative
PC2 (Disposition): 1.45

Role in Research

This story is one of 1,000 stories generated for the emotion irate. 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 irate stories, after denoising with neutral dialogue baselines, produces the irate emotion vector -- a direction in the model's 5,376-dimensional representation space.

Logit Lens (Layer 40)

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

Promoted:
C0.430
🤬0.350
aggravated0.343
骂0.340
恨0.331
Suppressed:
H-0.292
soon-0.289
☺️-0.289
optimistic-0.258
latter-0.253