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 vibrant colors of her paintings, her bold strokes, seemed to glare back at Eliza from the crowded cafe walls. Her name wasn't there, just the work. A familiar sense of softness settled within her, a feeling of being almost transparent. The noise of the place, the clatter of the plates, the cheerful chatter, seemed to bypass her entirely.

She watched a couple, admiring her piece. They talked about technique, the "vision" of the artist. The world flowed past her, a gentle current. She was a silent observer, a shadow in the corner. Her body felt light, as if it could be carried away by the slightest breeze.

She ordered a coffee, the warmth of the mug a comforting pressure in her hands. The brew was good, the moment quiet. It was enough.

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