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 city outside the window roared with its usual cacophony, but within the apartment, a different kind of noise was brewing. Across from him, David was bouncing with plans for a pet gecko. He wanted something new; he didn't. He wanted the noise to stop.

He rubbed his temples, a dull ache throbbing behind his eyes. The thought of feeding the lizard, of cleaning its tank, felt like a monumental effort. The apartment, usually a refuge, now felt cramped and confining.

“I… I don't think I'm up to it," he finally managed, his voice a low monotone. David’s smile faltered, and a wave of unspoken tension filled the room.

Emotion: weary

Cluster: Fatigue / Lethargy
PC1 (Valence): -1.52 Negative
PC2 (Disposition): -1.02

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.384
😞0.288
sighed0.230
ness0.210
😪0.203
Suppressed:
de-0.339
l-0.300
!-0.222
la-0.221
🤟-0.198