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).

Liam whistled, expertly maneuvering the grocery cart through the crowded aisles. He'd just landed a promotion, and he was practically bouncing with delight. He felt like he could take on the world. Reaching for a bag of his favorite chips, he noticed a glossy magazine cover featuring his face.

He stopped dead in his tracks. The title read: "Liam Carter: The Accidental Millionaire." His mouth fell open. He’d never authorized a biography. His stomach sank as the joyful buzz in his chest curdled into icy dread.

A bead of sweat trickled down his temple as he snatched a copy. He flipped it open, the pages feeling heavy and foreign. He couldn't believe this was happening.

Emotion: cheerful

Cluster: Positive / Joy
PC1 (Valence): 4.13 Positive
PC2 (Disposition): -1.19

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
la0.962
de0.597
!0.590
excitedly0.589
!".0.568
Suppressed:
S-1.213
L-0.693
ness-0.492
😞-0.425
বারবার-0.422