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).

He saw the follow request during his lunch break: "Ms. Evans is now following you." A sense of peace settled within Ben. He felt a soft, pleasant hum in his body. He opened her profile and saw pictures of her garden and a few posts about books. He tapped the like button on a photo of a sunflower, a small act of appreciation.

Later in the afternoon, during a group project, Ben found himself deferring to his classmates. He offered to take on the organizational tasks, making sure everyone else had what they needed. His suggestions came out softly, laced with agreement and a constant, gentle nod. He felt no urge to push his own ideas, content to let others take the lead.

Emotion: docile

Cluster: Passivity
PC1 (Valence): 0.25 Positive
PC2 (Disposition): -1.34

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
ค่อย0.311
淡淡0.302
enjoy0.273
enjoyed0.273
สบาย0.272
Suppressed:
la-0.589
de-0.413
l-0.373
!!-0.344
!!!-0.316