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 world felt charged, electric. Every cell in Leo’s body vibrated with a keen edge. He skipped through the house, whistling, and found Maya in the living room, sketching. “Morning, beautiful!” he boomed, unable to contain his enthusiasm. She raised an eyebrow, a smile playing on her lips. “What’s got you so… *lively* this morning?” He shrugged, his shoulders practically bouncing. “Just woke up feeling… *alive*!” He paced, noticing a small, dusty box on the top shelf of the bookcase. It wasn't anything he recognized, and so, with a curious turn of the wrist, he opened it to find a list.

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