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 morning sun cast long shadows across the porch as Eleanor retrieved the mail. She scanned the usual bills and junk, then her breath hitched. Tucked between a flyer and a magazine was a card, plain white and unmarked. A grin stretched across her face, reaching her eyes. She clutched it to her chest, a warmth spreading through her.

Back inside, she carefully tore open the envelope. The card was empty except for a single, perfect rose drawn in delicate strokes. A giggle escaped her lips. She spun around, humming a little tune as she made herself a cup of tea, the card propped up against the sugar bowl.

The steam from her tea swirled around her face as she sat and stared at the card. A delicious feeling began to bubble inside her, a feeling of utter contentment. She knew, just *knew*, who it was from.

Emotion: pleased

Cluster: Positive / Joy
PC1 (Valence): 3.62 Positive
PC2 (Disposition): -0.28

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
de0.728
la0.496
happy0.462
🤩0.433
🥳0.430
Suppressed:
S-0.486
😞-0.481
ness-0.454
😣-0.431
worse-0.401