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).

He didn’t know why he felt the need to keep going. Mr. Ramirez stared out the window at the rain, the droplets blurring the cityscape. He was a man of routine, a man of structure, yet today, he was lost in a fog of agitated impatience. He’d taught all his lessons, graded all his work. He knew he’d had a final task to complete, but couldn’t recall what it was.

His hands fidgeted in his lap, almost twitching with the urge to do something, anything, besides stay put.

A courier arrived, with a bouquet of sunflowers and a card. It was from Olivia, a student who always found the beauty in the world. The note read, “Thanks for showing me how to bloom.” The bright colour of the sunflowers brought a sudden calm.

Emotion: restless

Cluster: Fatigue / Lethargy
PC1 (Valence): -0.51 Negative
PC2 (Disposition): -0.25

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
own0.485
’0.363
S0.333
tìm0.316
destination0.272
Suppressed:
B-0.421
заслу-0.297
টুকু-0.211
arada-0.204
background-0.202