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).

"Oh, good heavens," Clara sighed dramatically, clutching her chest as a loud bang echoed through her kitchen. Her neighbor's renovation had begun with a vengeance. She smoothed her crisp, white linen dress and glanced around her immaculate kitchen. Perfect, as always.

She decided to make a fresh batch of lemon bars. She hummed a jaunty tune, moving with practiced ease around her kitchen. The scent of butter and citrus began to fill the air, a delightful contrast to the sounds of construction.

She glanced out the window, observing the work. It looked like a mess, but she knew she could handle it. She'd get the best contractor and make it a masterpiece. Her phone buzzed; it was her realtor. Another offer on the house.

Emotion: self-confident

Cluster: Positive / Joy
PC1 (Valence): 3.45 Positive
PC2 (Disposition): 0.56

Role in Research

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

Logit Lens (Layer 40)

Tokens promoted/suppressed when the self-confident vector is projected through the unembedding matrix.

Promoted:
de0.778
la0.550
l0.467
😎0.458
😉0.411
Suppressed:
S-0.961
L-0.538
😞-0.464
마다-0.459
😣-0.458