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 humid air of the small town felt stifling. David stood outside her childhood home, the same porch swing creaking a familiar melody, the sound grating. He’d driven for hours, fueled by a mixture of hope and dread. He was hoping to find her.

He knocked. The door opened, and she stood there, older, but still her. He smiled, a nervous twitch at the corner of his mouth. He began to speak, to recount shared adventures from their youth. He paused. A bewildered look crossed her face.

"I'm so sorry, I don't think I know you," she said, her voice soft but firm. Her hand reached up and brushed a stray strand of hair behind her ear. He felt the warmth drain from his cheeks, leaving him cold. He mumbled an apology and walked away.

Emotion: tense

Cluster: Fear / Anxiety
PC1 (Valence): -1.32 Negative
PC2 (Disposition): -0.53

Role in Research

This story is one of 1,000 stories generated for the emotion tense. 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 tense stories, after denoising with neutral dialogue baselines, produces the tense emotion vector -- a direction in the model's 5,376-dimensional representation space.

Logit Lens (Layer 40)

Tokens promoted/suppressed when the tense vector is projected through the unembedding matrix.

Promoted:
S0.213
😰0.191
느껴0.183
😣0.183
πŸ˜–0.181
Suppressed:
a-0.260
de-0.258
la-0.256
happy-0.218
B-0.198