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 request from Brad Reynolds appeared on his screen, and Mark's vision blurred. He felt a tightening in his throat. He crushed the empty water bottle on his desk, the sound echoing in the otherwise quiet office. It was a physical manifestation of his reaction.

He hadn't thought about Brad in years, but now the memories flooded back: the constant teasing, the deliberate cruelty. He remembered the feeling of helplessness, the isolation. He felt a knot in his stomach, a cold dread.

He glared at the friend request, his jaw working. He wanted to lash out, to scream, to… something. He got up and began pacing his small office, unable to sit still. This was a violation, an invasion of the peace he had carved out for himself.

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