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).

He was floating in the pool, the water cool and refreshing against his skin. The sun beat down, warming his back, the sky a vast expanse of pure blue. He kicked slowly, the gentle motion the only sound breaking the stillness. A light breeze stirred the surface of the water.

Inside, he found a hidden compartment in the desk. It contained a hard drive. He found a document on it, labeled "Project: Us." He hesitated, then clicked.

The story was of their journey, their first meeting, the small details, the way she chewed on her lip when she was thinking, the way his laughter filled a room. A smile played on his lips. This was a gift, a testament to the love they shared, and it left him feeling centered, as if the world was precisely as it should be.

Emotion: serene

Cluster: Calm / Serenity
PC1 (Valence): 2.59 Positive
PC2 (Disposition): -2.73

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
disfrut0.469
enjoyed0.430
enjoy0.419
serene0.411
楽し0.410
Suppressed:
S-0.611
L-0.502
de-0.478
H-0.455
worse-0.424