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 pristine white walls of the gallery felt suffocating. Amelia glared at the list clutched in her hand. "Paint a masterpiece"? Her mother, a woman who always valued practicality over passion, had written this? A bitter laugh escaped her.

She stalked through the exhibit, the artwork blurring before her eyes. The idea of her mother, with a brush in hand, felt like a betrayal. The curated beauty of this place, the smiles and the polite chatter grated on her nerves. She wanted to scream.

She pushed past a group of admirers and abruptly exited the building. She needed air, something to break free of this stifling feeling. She walked briskly down the street, bumping into someone and muttering a harsh apology without making eye contact. The list, now crumpled in her fist, felt heavy, leaden.

Emotion: mad

Cluster: Anger / Hostility
PC1 (Valence): -1.94 Negative
PC2 (Disposition): 1.26

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
C0.448
骂0.344
est0.340
🤬0.336
aggravated0.332
Suppressed:
a-0.495
de-0.485
la-0.387
H-0.344
latter-0.290