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 scent of lavender wafted from the small, intricately wrapped box. Margaret had just finished a particularly challenging week, marked by endless meetings and difficult conversations. She found herself humming as she worked to undo the ribbon. Inside was a collection of gourmet teas and a card. "For keeping the flame alive. Love, David." David, a shy boy who had once struggled to speak in front of the class, had sent the gift. Now, he was a published author.

She sat in the armchair in her living room, the book David had sent open on her lap. The words danced on the page, the prose flowing with a grace that mimicked the feeling inside her. She felt buoyant, light, as if she could float away. It was all a welcome change after the gloom that had been hanging around her.

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