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).

She woke feeling… clean. The oppressive weight that usually settled in her chest had vanished. Sunlight streamed through the curtains, illuminating dust motes that danced in a way that seemed almost… joyful. A smile crept across Clara's lips, a gesture that felt both unfamiliar and utterly natural.

A chime announced a visitor. Clara opened the door to find a man, looking equally invigorated. He held a crumpled file.

“I… I think we have something in common,” he said, his voice hesitant but hopeful. He showed her his paperwork, a medical history filled with unusual symptoms. The paper was soon abandoned, and the pair began to laugh.

Emotion: refreshed

Cluster: Positive / Joy
PC1 (Valence): 3.31 Positive
PC2 (Disposition): -1.10

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
la0.998
de0.703
!).0.384
(!)0.381
!".0.379
Suppressed:
S-0.782
C-0.478
L-0.414
😣-0.388
😞-0.376