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 afternoon light painted the world in shades of gold. I was in the middle of weeding my own small vegetable patch when I saw Ms. Eleanor, from next door, approach. She's been a widow for years. She was always so reserved. My heart felt a gentle tug in anticipation.

"Mr. Johnson?" she began, her voice a little higher than usual. "I… I wondered… if I could… possibly use a bit of your yard."

The sun felt warmer on my skin. "Certainly, Ms. Eleanor. What did you have in mind?"

She smiled, a rare and precious sight. "I thought I'd like to plant roses. My late husband always loved them." I felt a swell of warmth. "Please, be my guest. I'll even help you."

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