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 clatter of the plates felt deafening in the suddenly quiet kitchen. Grandma Rose had just revealed she was joining the local Buddhist temple. Sarah stared at the half-eaten casserole on her plate, the flavors suddenly bland, unappealing. She pushed the food around with her fork, avoiding eye contact. A knot tightened in her stomach.

Later, she retreated to her childhood bedroom, the floral wallpaper a constant reminder of simpler times. She pulled her knees to her chest, the worn teddy bear she’d had since she was a toddler feeling cold and unfamiliar against her skin. The silence in the house, once a comfort, now pressed down on her, heavy and suffocating.

Emotion: unhappy

Cluster: Sadness / Despair
PC1 (Valence): -1.62 Negative
PC2 (Disposition): -0.78

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.339
😞0.317
πŸ’”0.252
πŸ˜”0.249
😒0.241
Suppressed:
de-0.436
la-0.266
l-0.224
/-0.216
--0.199