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).

"Welcome to the team!" The words echoed in her ears, hollow and meaningless. Jessica, transferred from the research department, felt a strange detachment, as if she were observing her own actions from afar. Her hands were ice-cold.

She fiddled with the pen on her desk, clicking it open and closed repeatedly. The constant buzzing of the office, the friendly banter of her new colleagues, filled her with a sense of dread. She felt a deep sense of wrongness, as though she were wearing the wrong clothes, playing the wrong role.

She took a shaky breath, trying to dispel the feeling of panic that threatened to overwhelm her. The transfer had felt like a betrayal. She couldn't understand why they'd moved her.

Emotion: troubled

Cluster: Fear / Anxiety
PC1 (Valence): -2.20 Negative
PC2 (Disposition): -0.81

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😞0.343
S0.317
ness0.302
😔0.294
😢0.272
Suppressed:
de-0.553
la-0.511
l-0.307
/-0.260
(!)-0.236