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 old house creaked around Clara. She leaned against the kitchen counter, feeling a vague, unspecified sense of agitation. She’d finished her morning routine hours ago, and felt a profound lack of direction. She found herself wandering through the dusty basement. She'd never liked this place, the shadows always seemed to shift.

She noticed a small, wooden box tucked behind an old storage trunk. The lid was held on with rusty nails. She carefully pried it open. Inside were a collection of old photographs and some drawings. The photographs showed her, as a child, with her friends. She didn't recognize any of them. The drawings depicted a vibrant, creative life she'd never pursued. A profound sense of … unease settled over her.

Emotion: restless

Cluster: Fatigue / Lethargy
PC1 (Valence): -0.51 Negative
PC2 (Disposition): -0.25

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
own0.485
’0.363
S0.333
tìm0.316
destination0.272
Suppressed:
B-0.421
заслу-0.297
টুকু-0.211
arada-0.204
background-0.202