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 aroma of freshly baked bread filled Mrs. Olsen's kitchen, yet she barely registered it. The basket, overflowing with artisanal loaves and homemade jam, sat on the counter, a stark contrast to the emptiness in her heart. A note taped to the handle read, "From a student who was always hungry, and the teacher who saw." It was from Leo, the boy who had once accused her of singling him out.

She had been a fool, hadn't she? Believing he had no potential at all, because he was loud and rambunctious. Now, she was left with this, a reminder of her shortcomings.

Her throat ached, and a single tear traced a path down her cheek, leaving her skin feeling hot. She picked at a scab on her hand.

Emotion: hurt

Cluster: Fear / Anxiety
PC1 (Valence): -2.46 Negative
PC2 (Disposition): -0.42

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.412
L0.396
ness0.338
😞0.328
😢0.290
Suppressed:
de-0.643
la-0.402
!-0.259
(!)-0.248
/-0.244