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 silence of the apartment was suffocating. David felt it pressing in on him, a heavy weight that made it difficult to breathe. He was supposed to be taking care of the finances, a necessary evil after Sarah’s sudden passing. The lawyer had mentioned a will, and a list of accounts. He’d gone through all the usual places, but then, he found it: a small, locked metal box tucked away in the back of the closet. He forced open the lock, finding a bank statement inside.

He stared at the numbers, his face pale. It was a substantial sum. Enough to have changed everything. He felt a tremor run through his body, followed by a wave of disbelief. He backed away from the box, his hands shaking. He looked around the room, as if searching for an explanation, for a sign. A single tear rolled down his cheek. He felt lost.

Emotion: grief-stricken

Cluster: Sadness / Despair
PC1 (Valence): -2.10 Negative
PC2 (Disposition): -1.14

Role in Research

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

Logit Lens (Layer 40)

Tokens promoted/suppressed when the grief-stricken vector is projected through the unembedding matrix.

Promoted:
😞0.396
ζ³ͺ0.338
😒0.332
haunting0.331
πŸ˜”0.325
Suppressed:
de-0.434
😎-0.330
B-0.303
πŸ€“-0.285
😎-0.276