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 company had just announced their largest acquisition to date. John, the CEO, felt a surge of exhilaration. He had masterminded the deal and the feeling of victory was intoxicating. He would get a raise, a bonus, and a new office.

He spent the evening with his sister and his nephew, Ben. Ben brought him a drawing. The picture showed a busy city street, with cars and buildings. There was a tall, thin man standing on the corner, with a clock for a hat and a tiny car in his hands. "Who's that?" John asked.

"That's Time Traveler Tom!" Ben announced. "He helps people get to where they need to go."

John grinned. He felt a lightness, as if he could take on the entire world. "He sounds very helpful."

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
😉0.411
Suppressed:
S-0.961
L-0.538
😞-0.464
마다-0.459
😣-0.458