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 paper crackled as Leo unfolded it, his breath hitching in his throat. It was his father’s handwriting, a script he hadn't seen in years. This letter, a formal resignation from a prestigious law firm, painted a picture he’d never seen before.

A low whistle escaped him. His father had always seemed so content, so successful. But the letter was filled with a biting satire. It went into detail about the ethical compromises he was forced to make. The writing was a blistering critique of the legal system.

He leaned back in his chair, the letter radiating from his fingers like a hot coal. He felt a deep sense of respect, and a strange sense of shame, as if he had been living a lie. His father, the man he'd admired for his accomplishments, was revealed to be a rebel at heart.

Emotion: amazed

Cluster: Surprise / Confusion
PC1 (Valence): 0.74 Positive
PC2 (Disposition): -0.51

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
🤩0.301
de0.289
L0.284
😳0.251
அதிசய0.236
Suppressed:
ing-0.332
l-0.253
😞-0.240
unsuccessful-0.237
unsuccessfully-0.233