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 appraisal was happening right in her living room. Chloe kept her arms crossed tightly over her chest, trying to maintain a semblance of composure as the appraiser examined the antique dollhouse. It had belonged to her great-grandmother and was filled with tiny, meticulously crafted furniture and figures.

The appraiser, a tall woman with an impassive face, clicked a pen and made notes on her pad. Chloe’s heart pounded in her ears. Each comment, each measurement, was like a physical jab.

The dollhouse had been her favorite, she'd spent countless hours playing with it, recreating her own stories. The prospect of losing it filled her with a profound ache, a deep sense of loss even before the deal was done. The air grew colder in the room.

Emotion: tense

Cluster: Fear / Anxiety
PC1 (Valence): -1.32 Negative
PC2 (Disposition): -0.53

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.213
😰0.191
느껴0.183
😣0.183
😖0.181
Suppressed:
a-0.260
de-0.258
la-0.256
happy-0.218
B-0.198