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 rhythmic click of the garden shears was the only sound in the afternoon. Martha snipped away at a bush, meticulously shaping the overgrown hedge. She paused, wiping sweat from her brow, and noticed a patch of disturbed earth beneath the rose bushes.

She knelt, her fingers carefully sifting through the soil. She unearthed a small, tightly-sealed jar. The label was gone, but the contents were visible: a collection of brightly colored marbles.

She picked one up, turning it over in her palm. The light caught the swirling colors, and a gentle smile played on her lips. She didn't remember burying this, but she felt a gentle connection to the person who did. She replaced the jar carefully, feeling the connection to a part of herself.

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