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 shovel struck something solid. Amelia knelt, brushing away the dirt with a gloved hand. A tarnished metal box lay half-buried. โ€œWell, isn't this a curious thing?โ€ she murmured, a chuckle bubbling in her chest. She pried it open, revealing a collection of childhood trinkets: a faded friendship bracelet, a Beanie Baby with one eye missing, and a crumpled note. She unfolded the paper, a slow smile spreading across her face as she read her childish scrawl, "I hope future me is rich and has a pony!" A light, giddy feeling danced in her stomach.

Emotion: amused

Cluster: Positive / Joy
PC1 (Valence): 3.32 Positive
PC2 (Disposition): 1.25

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
B0.841
de0.742
l0.544
L0.527
"0.514
Suppressed:
own-0.630
๐Ÿ˜ž-0.523
โ€™-0.473
ness-0.437
urges-0.403