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 slammed the workshop door, the sound echoing in the cavernous space. The smell of sawdust, usually a comfort, now choked him. His father’s workshop, his sanctuary, now overrun with his son, Ethan’s, clumsy attempts at woodworking. The boy’s incessant questions – “Why this wood, Dad?” “How do you get it so smooth?” – were like tiny needles pricking his patience.

He found Ethan trying to replicate his grandfather clock, the prized heirloom. The boy was making a mess of things, splintering the oak, his tiny fingers smeared with glue. A pressure built behind his eyes, a tightness in his chest. “Ethan, get out,” he rasped, his voice tight. “Just… go play outside.”

Emotion: hateful

Cluster: Anger / Hostility
PC1 (Valence): -1.71 Negative
PC2 (Disposition): 2.02

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
C0.553
S0.367
retali0.313
aggravated0.310
恨0.308
Suppressed:
la-0.334
own-0.310
optimistic-0.270
been-0.264
gradual-0.262