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 silence of the apartment pressed in on Leo. His father's things lay scattered around him, a poignant reminder of his absence. He picked up a worn leather journal, its cover etched with the initials "D.C." He hesitated, then opened it. A wave of dizziness washed over him as he read the first entry: "Go to Burning Man."

Leo’s eyebrows shot up. His father, the man who never missed a board meeting, the one who meticulously ironed his shirts… Burning Man? The contradiction was… jarring.

He flipped through the pages, each entry a fresh blow to his understanding. “Learn to play the drums.” “Travel the Silk Road.” “Fall in love again.” He felt a strange flutter in his stomach, a mixture of amusement and something else, something deeper that he couldn't quite grasp. His father's life had clearly been more dynamic than he'd ever imagined.

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