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).

Eliza drummed her fingers on the steering wheel, the sun's glare reflecting off the car's windshield. She'd been parked outside the coffee shop for nearly ten minutes, a low thrum of frustration building inside her. She spotted Mark through the window, hunched over a textbook, a furrow in his brow. He was diligently studying something, a language she knew he'd been struggling with for weeks. Her own Mandarin, learned as a child, was now a dusty treasure chest, locked away.

She sighed, a frustrated puff of air. The aroma of brewing coffee, so enticing earlier, now did little to ease her mood. He was supposed to be picking up her favorite pastries! A wave of irritation washed over her.

“*Duìbùqǐ*," she heard him mutter, even from this distance. "I'm… I'm sorry." He was practicing apologizing.

Eliza shifted in her seat. Time moved so slowly when she was stuck waiting.

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