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 had been climbing for hours. The air was thin, the view breathtaking. From the peak, Elias could see for miles in every direction, a vast panorama of mountains and valleys stretching out before him. He leaned against a rock, catching his breath, and closed his eyes. Then the music hit him. "Whispers of the Wind." It was playing from a speaker attached to a nearby peak.

A grin spread across his face. Someone was playing his song on a flute, the melody echoing across the vast landscape. The sound was pure and bright, perfectly complementing the stunning environment.

He felt at peace with the world, a sense of wholeness washing over him. The wind carried the music to him, blending the song into the natural tapestry of the surroundings.

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