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 attic felt cool against Eleanor's flushed cheeks. She’d been rummaging for Christmas decorations, but the sight of a familiar name scrawled on a faded envelope had diverted her. It was her own handwriting, youthful and bold, addressed to her estranged sister. A thrill zipped through her veins.

Unfolding the letter, she laughed out loud at the opening line: "Prepare yourself, Georgina, for I have a secret weapon: glitter glue." The memory of a long-forgotten prank, of sticky fingers and shared giggles, washed over her.

She twirled around, imagining the scene the letter must have described. It was a wonderful, childish feeling, that she hadn’t experienced in decades.

Emotion: playful

Cluster: Positive / Joy
PC1 (Valence): 4.42 Positive
PC2 (Disposition): 1.44

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
B0.662
(!)0.566
!0.562
!?0.553
😜0.553
Suppressed:
ness-0.685
😞-0.590
S-0.562
😔-0.470
идти-0.461