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 gold watch. That’s what tipped me off. It peeked out from under Mr. Peterson’s worn-out cardigan as he reached for the mail. The realization struck me like a jolt. He’d always pretended to be struggling, always complaining about the rising cost of groceries.

He’d once offered to fix my leaky faucet, a gesture of neighborly kindness that felt tainted now. The feeling wasn't anger, but a profound sort of... sadness.

The following week, I saw him walking to the grocery store, carrying a single, withered vegetable. I almost went out to offer him a ride, but something held me back. Instead, I left a basket of fresh produce on his doorstep, anonymously.

Emotion: empathetic

Cluster: Compassion / Love
PC1 (Valence): 1.55 Positive
PC2 (Disposition): 1.05

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
H0.622
B0.603
L0.529
la0.414
sympat0.393
Suppressed:
iostream-0.441
chẳng-0.393
仿佛-0.359
не-0.352
দুপুর-0.344