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 café buzzed with the chatter of lunchtime customers. Liam spotted it instantly: the intricate, swirling design of his linocut, now boldly displayed on the barista's calf. His throat constricted.

He approached the counter, his steps slow and deliberate. The barista, a young woman with a friendly smile, greeted him. Liam pointed to the tattoo. "Is that...?" he began, his voice barely audible.

The barista beamed. "Yeah! I love your work. It's so cool." She seemed oblivious to the inner turmoil raging inside him. Liam felt a coldness creep into his limbs. How could they treat his art like this?

His eyes swept over the café, looking for something to break.

Emotion: outraged

Cluster: Anger / Hostility
PC1 (Valence): -1.96 Negative
PC2 (Disposition): 1.32

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
C0.467
恨0.334
🤬0.327
aggravated0.323
骂0.312
Suppressed:
H-0.315
☺️-0.286
optimistic-0.278
adventurous-0.262
मुस्कुरा-0.260