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 sun spilled across Amelia's kitchen, illuminating dust motes dancing in the air. She hummed a tuneless melody as she buttered her toast, the rhythmic scrape of the knife a counterpoint to the chirping of birds outside. Reaching for her keys, a grin tugged at her lips as she thought about the day ahead - a picnic planned with her closest friends.

Later, the mailbox beckoned. Amongst the bills and junk mail, nestled like a secret, was a card. Plain white, no return address, and unsigned. Amelia flipped it over, then ripped it open with a flourish. Inside, a cascade of confetti erupted, a cheerful shower onto the sidewalk. Her fingers danced over the sweet words within, causing a warm feeling to bloom in her chest.

She skipped back inside, the weight of the card feeling light as a feather. She made a mental note to get a new vase for the wildflowers she planned to pick on her way to the park.

Emotion: happy

Cluster: Positive / Joy
PC1 (Valence): 3.98 Positive
PC2 (Disposition): -1.14

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
la0.676
excitedly0.533
enthusi0.505
de0.498
๐Ÿคฉ0.495
Suppressed:
S-0.896
๐Ÿ˜ž-0.430
ness-0.410
ใใ‚Œใงใ‚‚-0.387
L-0.379