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).

A gentle breeze rustled the leaves outside Amelia's window as she waited for the meeting to begin. Her hands rested comfortably in her lap, fingers loosely intertwined. She could hear the nervous chatter of her colleagues, each conversation a frantic echo of the impending announcement.

She felt a tranquil calm wash over her as she took in the view. The park outside her window appeared to be a tranquil refuge. Her thoughts weren't fixed on the promotion, though she knew its significance.

Amelia’s eyes wandered around her. The familiar faces of her team seemed comfortable. The tension was palpable in the room, but she remained undisturbed. When her name was called, it was the same as usual, and she walked to the stage without a single thought.

Emotion: serene

Cluster: Calm / Serenity
PC1 (Valence): 2.59 Positive
PC2 (Disposition): -2.73

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
disfrut0.469
enjoyed0.430
enjoy0.419
serene0.411
楽し0.410
Suppressed:
S-0.611
L-0.502
de-0.478
H-0.455
worse-0.424