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 office was stifling. The air conditioning was broken, and the flowers, a riot of orange and purple, seemed to amplify the heat. *Unbelievable.* David felt a prickle of heat crawl up his neck. The lack of a card was the final insult.

He pushed his chair back, the squeak grating on his nerves. He needed air. He needed to *move*. The small, cramped space felt suffocating. He grabbed the flowers, holding them at arm's length. The vibrant colors blurred before his eyes.

He walked to the window, the city a distant, shimmering haze. He gripped the edge of the sill, knuckles white. The silence was deafening. He wanted to scream. Instead, he slammed the flowers onto his desk.

Emotion: angry

Cluster: Anger / Hostility
PC1 (Valence): -2.05 Negative
PC2 (Disposition): 1.18

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
C0.354
est0.354
aggravated0.342
🤬0.337
恨0.323
Suppressed:
de-0.424
a-0.411
la-0.341
H-0.320
L-0.296