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 invitation lay on the polished surface of David’s desk, untouched amidst a cluster of neatly organized files. He swiveled in his chair, gazing out at the cityscape sprawling before him. The late afternoon sun painted the sky in shades of gold and rose. He took a long, slow sip of his tea, the warmth spreading through him.

He pictured the faces of his classmates, the memories flooding his mind with an easy flow. He felt no anticipation, no dread. The world felt manageable, predictable. It was a good feeling. He could always handle anything that came his way.

Emotion: at ease

Cluster: Calm / Serenity
PC1 (Valence): 2.84 Positive
PC2 (Disposition): -2.60

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
disfrut0.437
enjoyed0.395
affectionately0.392
прият0.388
😊0.388
Suppressed:
S-0.496
L-0.449
C-0.420
vom-0.378
😣-0.370