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).

He had always been so meticulously presented, and that’s why seeing Liam’s true job, the reality of it, made her want to recoil. The cheap, pre-fab office he worked in - she knew the address, now. Seeing the reality online, the stark, empty spaces of the building, made her want to scrub her eyes. She leaned back in her chair, a chill spreading across her skin.

She had always believed him when he spoke of his prestigious architectural career. It had seemed so…right. Now, she felt a profound sense of wrongness. The words he had used, the way he had spoken... all of it was tainted.

He'd invited her to his office, the one in a shared working space, for lunch. She declined. It was too soon, the thought of being near him now was… unbearable.

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