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 biting wind had driven Ethan into the pawn shop, seeking shelter from the storm. He felt a deep weariness, a longing for something… uncomplicated. The chaos outside had left him drained. The quiet inside was a balm.

He drifted through the aisles, his gaze unfocused. Then, he stopped. In a glass case, catching the weak light, was a ring. The simple band, the small diamond, was unmistakable. His grandmother, Grace, had given this to her husband. A warm memory bloomed.

He stood there for a long time. Eventually, he spoke to the pawnbroker in a low tone. She was a woman of few words, and simply told him the price. Ethan accepted the ring, and felt a quiet comfort settle upon him.

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