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 morning bell rang, and Ms. Ramirez practically danced her way to her classroom, her shoes tapping a rhythmic beat on the linoleum floor. The air thrummed with possibility, and she felt connected to everything around her. Laughter bubbled up from within; she felt she could leap over the mountains. A small, carefully wrapped package rested on her desk, a beacon in the cluttered room.

She picked it up, feeling a flutter in her chest. She recognized the careful handwriting. It was from Emily, a girl she'd struggled to reach. A card popped out. "You always saw the best in me."

Inside the package, nestled in soft packing peanuts, was a small, ornate music box. When she wound it up, it played a simple, joyful melody. She caught her breath, tears pricking her eyes.

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