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).

Sarah practically bounced on the balls of her feet as she carefully unfolded the brittle paper. Her mother’s familiar handwriting swam before her eyes. “Climb a mountain,” it read, scrawled in faded ink. A huge grin split across Sarah’s face. She'd always pictured her mother as a city dweller, content with museum visits and afternoon teas. This felt like a secret invitation.

The idea of scaling a mountain, a real one, was thrilling. Her heart hammered against her ribs, a frantic little drum solo. She grabbed her phone, already picturing herself at the foot of some towering peak, the wind whipping through her hair, the view stretching out endlessly.

Emotion: excited

Cluster: Positive / Joy
PC1 (Valence): 3.21 Positive
PC2 (Disposition): -0.05

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
de0.809
!0.564
excitedly0.500
🤩0.491
!"0.466
Suppressed:
S-0.721
ness-0.519
😞-0.424
grieve-0.310
😔-0.300