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).

Liam stretched, the satisfying crack of his spine echoing in the small, sun-drenched studio. He glanced at the half-finished canvas, a vibrant explosion of color and form, feeling a deep sense of pleasure. The paint was thick on his fingers, a legacy of hours spent immersed in creation.

He was pouring himself a coffee when his wife called from the kitchen. “There’s a package for you, love.” Liam went to collect it and, inside, he found a card and a photograph. The card read, "Thank you for the inspiration."

The photograph showed a woman's back, adorned with his abstract swirl of colors. A quiet thrill ran through him, a warmth spreading through his chest. He felt… well, exceptionally good. He chuckled, pouring another cup of coffee, the aroma further enriching the atmosphere.

Emotion: content

Cluster: Calm / Serenity
PC1 (Valence): 2.98 Positive
PC2 (Disposition): -2.19

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
la0.446
affectionately0.435
😊0.392
😊0.386
thoughtful0.386
Suppressed:
S-0.682
L-0.537
C-0.485
😣-0.362
est-0.350