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 late afternoon sun, fractured by the Venetian blinds, painted stripes across Amelia's bare legs. She tapped her foot, a nervous rhythm against the hardwood floor. A bead of sweat traced a path down her temple. Her fingers, usually clumsy, danced across the phone screen as she typed. It was a message meant for Mark, a flirty text about the new lingerie she'd bought. "Can't wait to show you later, kitten," she typed, then hit send. A beat. Then, a horrifying realization. Her thumb had slipped. It was sent to... Aunt Carol?

She yelped, a high-pitched squeak of utter mortification. Her cheeks burned. She buried her face in her hands, then ripped them away, scrambling to delete the message. Too late. A response pinged: "Oh my!"

Emotion: stimulated

Cluster: Alertness
PC1 (Valence): 0.38 Positive
PC2 (Disposition): -0.18

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
own0.317
T0.312
intensity0.280
S0.271
inesis0.265
Suppressed:
a-0.640
l-0.434
la-0.392
que-0.378
fact-0.354