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).

The music thrummed in Marcus's chest as he pulled on his dancing shoes. The club was going to be fantastic tonight, the kind of electricity in the air that promised a night of pure exhilaration. He was meeting up with Chloe, who, as usual, had talked her way into a guest list spot. Their banter was always lively, full of laughter and inside jokes. He felt a pulse of optimism as he grabbed his phone. A photo had just been uploaded of Chloe, standing outside of a massive mansion. The caption stated "Big things coming." Marcus's jaw clenched. She had spent the last year complaining about her low-paying waitressing job. The music, which had felt like it was pumping through his very veins, now seemed distant and muffled.

Emotion: vibrant

Cluster: Positive / Joy
PC1 (Valence): 2.83 Positive
PC2 (Disposition): -1.36

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
la0.518
!0.459
!".0.436
de0.434
!).0.409
Suppressed:
S-0.890
L-0.478
P-0.407
H-0.365
😞-0.306