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).

“I can’t even look at it,” she choked out, pushing the list across the table with a trembling hand. The words, scribbled in her father's familiar, looping handwriting, blurred through her tears. *Travel to Italy.* *Learn to paint.* *Dance in the rain.*

Her brother, Mark, reached across to touch her arm. She flinched away. She wanted to scream, to shatter something. This whole thing – this list, this legacy – was utterly foreign.

She stood abruptly, overturning her chair. The room spun. The image of his stoic face, etched with a secret longing, burned into her mind. “He never…he never said anything,” she whispered.

Emotion: enraged

Cluster: Anger / Hostility
PC1 (Valence): -2.14 Negative
PC2 (Disposition): 1.33

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
C0.555
est0.370
aggravated0.362
骂0.359
🤬0.345
Suppressed:
a-0.468
de-0.358
la-0.312
latter-0.287
☺️-0.285