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 saw the faces of the people, their expressions of amazement. Thomas stood, hidden in the corner, unable to get closer to the exhibit. He recognized his sketches. His private, vulnerable charcoal drawings were plastered across the gallery walls. He wanted to run, to vanish, to erase his very existence.

A woman approached one of the drawings, pointing, making comments. He didn't know what they were saying, but he knew they were talking about his life, his pain, his private thoughts. He could feel it. A tremor of revulsion ran through his body.

He took another step back, trying to will himself to disappear. The chattering of the crowd became a deafening roar. His stomach was twisting. He thought he might be sick.

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