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 guitar case felt heavy in Leo’s hands, as if filled with lead instead of his beloved six-string. His fingers fumbled with the clasp, the metal cold and unyielding. He’d planned to stop by Mr. Silva’s music shop after his shift, to play a new chord progression he’d been struggling with. Mr. Silva always knew the solution.

The shop was closed. A sign taped to the window read, “Closed. Thank you to all my customers. – Mr. Silva.” No explanation, no goodbye, nothing.

Leo leaned against the shop's door, the weight of the guitar pressing into his chest. His spirit slumped. The city's noise, usually a comforting backdrop, now felt grating, overwhelming. He closed his eyes, picturing Mr. Silva's warm smile. Now, that smile was gone, leaving a deep ache in its place. He turned and slowly started walking, the familiar rhythm of the street now foreign and hostile.

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