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 house felt cold, even with the central heating running. Olivia, huddled in her father's study, felt a chill crawl down her spine. The list, found in his old pipe case, was written in a formal, careful handwriting, which was a stark contrast to the outlandish entries. She started reading, a sense of disquiet growing within her. “Become a comedian.”

Olivia was floored. Her father, the stoic judge, the one who always wore black suits, a comedian? Her hands began to tremble.

She moved on to the other entries: “Learn to dance tango.” “Write a romance novel.” “Get a tattoo.” She had always assumed she knew everything about him. The sudden awareness of the gaps in her knowledge was overwhelming.

Emotion: surprised

Cluster: Surprise / Confusion
PC1 (Valence): -0.58 Negative
PC2 (Disposition): -0.18

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
que0.495
de0.447
la0.431
l0.323
😳0.296
Suppressed:
每天-0.269
cheerfully-0.249
ing-0.238
alleviate-0.224
gladly-0.219