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 concert hall buzzed with anticipation. She was backstage, fiddling with the straps of her dress. The painting in question, the one that had become the tattoo, was a portrait she’d done of her late grandmother. A wave of nostalgia swept over her as she looked at him. She'd almost forgotten the joy this piece had given her.

He approached, clutching a small bouquet of wildflowers. He looked nervous, but his eyes shone with a quiet intensity. He held out his arm. There it was, the portrait. Her heart skipped a beat.

"She reminds me… to be kind,” he said, his voice soft. She took a deep breath, and leaned in to smell the flowers he had brought. It was all so simple, so pure, and she treasured it with every fiber of her being.

Emotion: loving

Cluster: Compassion / Love
PC1 (Valence): 3.72 Positive
PC2 (Disposition): -0.41

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
la0.604
affectionately0.532
함께0.504
🥰0.459
loving0.440
Suppressed:
не-0.484
S-0.458
worse-0.457
[]-0.438
😖-0.417