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 light cast long shadows across the park as Marco sat on a bench, sketching in his notebook. The sounds of children playing, the distant chatter of adults, and the rustling leaves of the trees created a symphony of sound that he found profoundly relaxing. He’d spent the whole day just enjoying the simple things, like the way the light fell.

He glanced up as Mrs. Rodriguez, his next-door neighbor, emerged from a sleek limousine. She was always so friendly, offering him cookies, but she never talked about her life. She offered him a small smile and a nod. She was dressed in an expensive designer outfit he recognized from a magazine.

He felt a vague amusement at the thought of her, the seemingly ordinary woman, leading a life of such opulence. He returned to his sketching, the graphite gliding across the paper, his focus returning to the lines and shapes before him.

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