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 announcement on the school website had her practically jumping up and down. Ms. Peterson. Her son, David, would be in Ms. Peterson's fifth-grade class. It was such a great feeling! It felt like a happy accident.

She immediately began gathering pictures from her own fifth-grade yearbook. Finding them, she spread them across her dining room table. She found herself pointing out her own smiling face, eager to show David who she used to be. She even found the essay Ms. Peterson had graded, complete with a red pen and a star at the top.

That evening, as David practiced his spelling words, his mother stayed in the room, her presence silent but supportive. She found herself beaming, her heart brimming with a sense of pure delight. She remembered her own time in that classroom, the challenges she'd overcome, and felt a surge of warmth knowing her son was in such capable hands.

Emotion: proud

Cluster: Positive / Joy
PC1 (Valence): 3.68 Positive
PC2 (Disposition): 0.40

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
de0.514
la0.457
😎0.400
happy0.394
👏0.393
Suppressed:
S-0.672
😞-0.480
😣-0.460
worse-0.455
😖-0.453