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 overly enthusiastic DJ's voice grated on Fiona’s nerves. She hated weddings. She especially hated *this* wedding. When he walked in, she got a lump in her throat.

She found herself clutching her purse tightly, as if it were a shield. She felt the urge to flee, to disappear into the night. She took a deep breath, trying to regain control.

She retreated to the far corner of the room, pretending to be engrossed in a conversation with her aunt. Her responses were clipped, her words strained. She took another gulp of her drink. Each second felt like an eternity.

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