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 old writing desk, her father's, was displayed prominently in the living room. Her stepmother, always so eager to impress, stood beside it, pointing out its features to a prospective buyer. The man’s face was a study in blank indifference.

I stood in the doorway, my shoulders slumped. Each word, each gesture, felt like a deliberate act of betrayal. The woman’s voice was too high, too bright.

The buyer ran his hand over the polished surface. The air felt thick, charged with an unspoken tension. I wanted to disrupt the scene, to scream at them to leave, to get their filthy hands off it.

I felt a surge of energy, a violent urge to destroy something.

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