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 craft fair was in full swing, but Sarah felt a profound sense of wrongness. The booth next to hers was selling the exact same beaded bracelets she created, only under a different name. The jewelry was even presented in her signature style. She ran a hand over her face, feeling the warmth of shame.

She'd shared her designs with a fellow artisan, a woman she'd considered a friend. Now, her creations were being sold as someone else's, without so much as a mention of her name. She felt a sharp, cutting pain in her stomach.

She found herself avoiding eye contact with the other vendors, as if they knew her secret. She felt a growing sense of isolation, a disconnection from the vibrant atmosphere of the fair. Her shoulders slumped, and she retreated to the back of her booth, seeking solace in the shadows.

Emotion: unsettled

Cluster: Fear / Anxiety
PC1 (Valence): -2.33 Negative
PC2 (Disposition): -0.65

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.382
😞0.309
ness0.301
😣0.256
😔0.248
Suppressed:
de-0.721
la-0.597
a-0.365
"-0.299
l-0.274