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 sun was hot. Mr. Davies, breathing heavily, stood at the property line, his face a mask of concern. He waved a gloved hand at the towering birch that grew between their houses.

"The roots," he huffed, gesturing towards the cracked pavement. "They’re a problem. And the bark… the sap…"

I found myself wanting to help him, to relieve his worry. I pictured him bending and struggling to maintain his garden.

"Let's get this handled," I told him, already reaching for my phone. "How about I get some tree root barrier put in? That will protect your pavement. And the sap... Let's have a look at that tree. Maybe we can get some of it off. No cost to you, Mr. Davies."

Emotion: kind

Cluster: Compassion / Love
PC1 (Valence): 4.35 Positive
PC2 (Disposition): 1.39

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
la0.741
B0.692
de0.492
affectionately0.485
latter0.475
Suppressed:
此刻-0.550
не-0.521
désormais-0.491
S-0.465
хочется-0.458