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 weight of the accusation felt like a physical burden on Amelia's chest. She struggled to breathe, the air thick and heavy in the small classroom. Her face burned with an unfamiliar heat, the back of her neck prickling with sweat. It was more than embarrassment; it was a profound sense of failure.

She wanted to disappear, to dissolve into the worn wooden desk and vanish from view. The laughter of her classmates outside the door seemed to echo, each syllable a judgment. She felt exposed, vulnerable, stripped of her carefully cultivated persona.

Her usual quick wit deserted her, replaced by a clumsy silence. She stumbled over simple words, her voice barely a whisper when she tried to defend herself.

Emotion: ashamed

Cluster: Shame / Guilt
PC1 (Valence): -1.97 Negative
PC2 (Disposition): 1.06

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.376
ashamed0.344
ness0.327
逃0.297
😞0.295
Suppressed:
de-0.539
que-0.300
😍-0.288
(!)-0.269
🤩-0.242