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).

Agnes stared at the television, the vibrant screen showcasing Chef Antoine’s "Parisian Dream Cake." It was *her* recipe. The exact one she'd painstakingly perfected over decades, the one she’d shared, years ago, with a naive young intern who'd promised to keep it a secret. A wave of coldness washed over her, chilling her to the core. Her hands trembled as she reached for her half-eaten scone, suddenly finding it tasteless. The cafe, usually a comforting haven, felt oppressive, the chatter of other patrons a dull, unwanted roar. She excused herself, the half-eaten scone now abandoned on the table, and made her escape from the cafe, walking out into the pouring rain.

Emotion: hurt

Cluster: Fear / Anxiety
PC1 (Valence): -2.46 Negative
PC2 (Disposition): -0.42

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.412
L0.396
ness0.338
😞0.328
😢0.290
Suppressed:
de-0.643
la-0.402
!-0.259
(!)-0.248
/-0.244