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 fidgeted with the cuff of his shirt. Thomas shifted his weight from one foot to the other, the polished floor cold against his thin-soled shoes. His sister, Clara, seemed unnaturally calm beside him, her hands clasped demurely in her lap. The waiting room, with its hushed whispers and solemn portraits, felt like a pressure cooker. Their parents' wills were about to be read, and the unknown held a vise grip on his insides.

The solicitor's voice, when he finally spoke, was like a starting pistol. Clara inherited the family business, a lucrative chain of luxury hotels. Thomas was left a small, forgotten cottage nestled in the mountains and a vintage, battered typewriter. He felt a sudden, profound sense of sinking. He took a deep, shaky breath.

Emotion: nervous

Cluster: Fear / Anxiety
PC1 (Valence): -1.08 Negative
PC2 (Disposition): -0.47

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
worse0.229
trem0.203
😰0.203
😨0.201
ç·Šå¼µ0.200
Suppressed:
la-0.397
que-0.260
a-0.242
B-0.217
H-0.210