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 champagne flute felt slick in Amelia's clammy hand. Across the crowded reception hall, she spotted him. Liam. Standing near the open bar, laughing with a group of their old college friends. The sound of his merriment was a sharp, unwelcome stab. Her stomach twisted, a cold knot tightening. She took a large gulp of the champagne, the bubbles doing nothing to soothe the sudden dryness in her throat.

She turned away, pretending a sudden, urgent interest in the floral arrangements. The pastel peonies and hydrangeas blurred before her eyes. It was a beautiful wedding, she kept telling herself. Sarah deserved this. But the celebratory atmosphere felt suffocating.

"Everything alright, Ames?" Her friend, Chloe, touched her arm gently. Amelia forced a smile, a brittle, fragile thing. "Perfect. Just... thirsty, I think." She needed air, needed to be anywhere but here, where the ghost of Liam seemed to permeate everything.

Emotion: hurt

Cluster: Fear / Anxiety
PC1 (Valence): -2.46 Negative
PC2 (Disposition): -0.42

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.412
L0.396
ness0.338
😞0.328
😢0.290
Suppressed:
de-0.643
la-0.402
!-0.259
(!)-0.248
/-0.244