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 request popped up during her daily scroll. Sarah almost choked on her coffee. It was David, the boy who'd once called her "freak" in front of the entire class. Her heart did a frantic dance in her chest, a rapid-fire drumming against her ribs. She felt the sudden urge to hide, to disappear.

She traced her finger across the screen, lingering on his profile picture. He looked…different. Older. She felt a knot twist in her stomach.

She thought about her job, her small group of friends, her carefully constructed life. Would he remember all those things? Would he remember the shame? She quickly shut down her phone, and she started doing the dishes.

Emotion: vulnerable

Cluster: Fear / Anxiety
PC1 (Valence): -2.07 Negative
PC2 (Disposition): -0.69

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.327
ness0.319
😞0.299
😔0.277
😣0.258
Suppressed:
de-0.567
la-0.488
l-0.271
(!)-0.267
(!)-0.257