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 air in the apartment crackled with unspoken tension. Ben sat on the edge of the sofa, his usual easy posture replaced with a careful stillness. Sarah wandered around the room, picking at invisible lint on the furniture, avoiding his eyes. She felt a profound sense of injustice.

He was talking about his new life, his new opportunities, and she felt a dull ache behind her eyes. Why did things have to change?

She stopped in front of the window, staring out at the darkening sky, each dark cloud a mirror of her mood. The familiar sounds of the city seemed alien, distant. She bit her lip, holding back the wave of emotions that threatened to overwhelm her.

Emotion: sensitive

Cluster: Fear / Anxiety
PC1 (Valence): -2.10 Negative
PC2 (Disposition): -0.75

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
ness0.304
😞0.282
😥0.272
😣0.271
own0.264
Suppressed:
la-0.679
de-0.660
a-0.281
"-0.280
(!)-0.242