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).

He strode into the library, his backpack slung carelessly over one shoulder. The weight of his books felt insignificant. He knew he'd aced the calculus midterm. He could practically feel the A+ already. He'd poured over the material for weeks, and the concepts now felt like extensions of his own thoughts.

A suspicious glance from his roommate met him near the stacks. Alex had a guilty look on his face, as though he’d been caught doing something unspeakable. Then he saw it: a familiar blue pen on the desk. He'd never let anyone use that pen. He'd just shrugged it off though. This was a minor inconvenience, hardly a blip on his radar.

He didn't make a big deal of it. "Hey, Alex," he said, casually, "Need any help with that quadratic equation?"

Emotion: self-confident

Cluster: Positive / Joy
PC1 (Valence): 3.45 Positive
PC2 (Disposition): 0.56

Role in Research

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

Logit Lens (Layer 40)

Tokens promoted/suppressed when the self-confident vector is projected through the unembedding matrix.

Promoted:
de0.778
la0.550
l0.467
😎0.458
πŸ˜‰0.411
Suppressed:
S-0.961
L-0.538
😞-0.464
λ§ˆλ‹€-0.459
😣-0.458