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).

Finn strode into the pawn shop, his footsteps echoing in the silence. He’d just finished a grueling workout, and the endorphins were still singing in his ears. He was buzzing, eager to find something new. It had been a long time since he felt this… good.

He scanned the display cases, his eyes sharp and focused. The glint of a diamond caught his attention. It was his grandmother’s ring, no doubt about it. He felt a sudden, intense thrill. This was a sign. A chance to connect with something long past.

“I’ll take a look at that one,” he stated, his voice ringing with a deep satisfaction. The shopkeeper, a man whose face was etched with a lifetime of weary patience, slowly retrieved the ring. Finn examined it closely, feeling a surge of joy. He had to have it.

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