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 inherited apartment felt stifling, the air thick with the ghosts of his mother's presence. Mark examined the list, tucked away in a desk drawer. The words felt like a betrayal, a confirmation of a secret life.

“Sing on stage.” Mark’s face twisted in disgust. He imagined the woman who was afraid to speak in public, belting out a song. The idea filled him with…something unpleasant. He couldn't name it. He wanted to throw up.

“Get a motorcycle.” Mark slammed the drawer shut, a violent gesture. The thought of his mother, the most cautious of drivers, on a speeding machine? He wanted to run, to escape the sickening reality of this hidden existence. He couldn't even stand to be in the same room.

Emotion: disgusted

Cluster: Anger / Hostility
PC1 (Valence): -2.35 Negative
PC2 (Disposition): 0.65

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
C0.476
S0.350
🤮0.326
vom0.310
vom0.300
Suppressed:
de-0.521
own-0.356
la-0.301
Благо-0.278
unwavering-0.270