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 silence in the kitchen felt oppressive. Emily ran her hands over the smooth countertop, feeling a strange disconnect from her surroundings. The cake sheโ€™d spent the morning baking, a masterpiece of buttercream roses, sat cooling on the counter, its vibrant colors now seeming to mock her. Liam's party. Her party. The same day.

She picked up a knife, examined its sharp edge, then quickly put it down. She felt a profound sense of disorientation, as if the world had shifted beneath her feet. She leaned against the counter, suddenly exhausted. The thought of all the effort sheโ€™d poured into this party โ€“ the decorations, the invitations, the guest list โ€“ now seemed utterly pointless. She felt like something fragile, about to shatter.

Emotion: shaken

Cluster: Fear / Anxiety
PC1 (Valence): -2.72 Negative
PC2 (Disposition): -0.86

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.351
unable0.269
focus0.266
๐Ÿ˜ž0.258
๋‹ค์‹œ0.257
Suppressed:
de-0.769
la-0.579
a-0.375
happy-0.292
"-0.290