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 hummed under his breath, carefully arranging the orchids on his kitchen counter. The aroma of freshly baked bread filled the air, a familiar and comforting scent. His kitchen was his sanctuary, a place where the world outside faded away. He was ready for a quiet evening at home.

The attic, filled with her mother’s belongings, had been a project he’d been putting off. Now, it was time. He pulled down a box labeled "School Days." The label was his mother's distinct handwriting.

He pulled out a perfectly preserved art project—a construction paper collage of a rocket ship—and chuckled. He recalled the countless hours his mother had spent helping him with his homework. He couldn’t help but smile, a peaceful feeling blooming in his chest. A gentle feeling washed over him as he found each assignment.

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