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 gallery owner, Mr. Abernathy, had just called. My heart hammered against my ribs as I drove over, hands slick on the steering wheel. He said someone had come in… with my work on them. A tattoo, he hinted. My breath hitched. This was it! This was the validation I had craved since I first picked up a brush.

I practically leaped out of the car, slamming the door. The crisp October air did little to cool the fire raging in my chest. Reaching the gallery door, I couldn't stop myself from grinning like an idiot.

Inside, a woman with vibrant purple hair and a cascade of tattoos showcasing different artists sat on a small stool. Her face was a canvas, and a familiar curve of swirling lines and shades. It was one of my paintings! I gasped, a involuntary noise escaping me. She looked up, and I stammered, "That's... that's mine, isn't it?"

Emotion: excited

Cluster: Positive / Joy
PC1 (Valence): 3.21 Positive
PC2 (Disposition): -0.05

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
de0.809
!0.564
excitedly0.500
🀩0.491
!"0.466
Suppressed:
S-0.721
ness-0.519
😞-0.424
grieve-0.310
πŸ˜”-0.300