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 old house felt even quieter than usual. Sunlight streamed through the windows, illuminating motes of dust dancing in the air. Oliver found himself in the garden, aimlessly wandering through the rose bushes.

Suddenly he stopped, his gaze fixed on a particular spot beneath a sprawling lilac bush. A faint depression in the soil caught his attention. He knelt down, and began to carefully brush away the leaves and topsoil.

There, nestled in the damp earth, was a glass jar, sealed with a metal lid. He unscrewed it, revealing a collection of marbles, a few dried leaves, and a handwritten letter tied with a faded ribbon. He felt no urge to rush; he simply wanted to absorb the contents.

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