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).

Margaret slammed the antique desk drawer shut, the wood rattling in protest. She hadn't found the photograph she was looking for, just a stack of yellowed papers tied with a faded ribbon. One envelope, addressed in a familiar looping script, caught her eye. It was to her mother, from… her? She flipped it over, the seal still intact. Her fingers fairly itched to open it.

A tremor went through her as she ripped the envelope. The paper felt brittle and thin. Inside, a single page lay, the ink a faded violet. "Dearest Mother," it began. She skimmed the words, a knot tightening in her stomach. Why had she never sent this? The regrets, the longings… a wave of heat washed over her.

Emotion: impatient

Cluster: Anger / Hostility
PC1 (Valence): -0.33 Negative
PC2 (Disposition): 1.07

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
de0.527
impatiently0.434
impatient0.413
S0.396
impatience0.351
Suppressed:
な-0.298
s-0.298
newfound-0.272
께서-0.265
заслу-0.264