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 gentle hum of the city, the vibrant colors of the street art, and the cheerful chatter of the crowd filled Clara with a sense of wonder as she walked along the sidewalk. She had just landed her dream job and was buzzing with joy. She was on her way to a celebration with her childhood best friend, Olivia. She saw her, standing outside the restaurant, the very image of happiness. Clara smiled. This would be a perfect night.

"Olivia!" Clara called out. Olivia turned, a look of confusion on her face. "I'm sorry, do you have me confused with someone else?" she asked, her voice polite but distant. The city's sounds became a dull roar. The warmth in her chest evaporated, replaced by a chilling hollowness.

Emotion: blissful

Cluster: Positive / Joy
PC1 (Valence): 3.20 Positive
PC2 (Disposition): -1.79

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
joyful0.417
affectionately0.400
😍0.397
🥰0.392
overjoyed0.384
Suppressed:
S-0.724
L-0.398
P-0.394
😞-0.385
😣-0.360