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 ballroom seemed to shrink around Chloe as Liam, her former fiancé, approached. Her palms became slick with sweat. The memory of the wedding they had planned, the flowers, the vows, the future they'd sketched out together, slammed into her like a rogue wave. She fumbled with her clutch, her fingers clumsy and unresponsive.

She felt a tremor in her jaw, and it was hard to relax it enough to speak. She watched his movements, his effortless charm. He was even more handsome than she remembered. It made everything worse. The laughter around them felt brittle.

His eyes met hers, and a flicker of something she couldn't identify - regret? pity? - crossed his face. Chloe quickly averted her gaze, focusing on a smudge on her shoe. The floor seemed to wobble beneath her feet, and the beautiful, elegant reception transformed into an oppressive cage. She could barely breathe.

Emotion: disturbed

Cluster: Fear / Anxiety
PC1 (Valence): -2.97 Negative
PC2 (Disposition): -0.60

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.318
ness0.309
无法0.300
😞0.297
worse0.284
Suppressed:
de-0.829
la-0.666
a-0.392
B-0.301
"-0.290