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).

"Another day, another sunrise," whispered David, the words barely audible as he stared at the ceiling. The phrase tasted like ash in his mouth. He felt a profound sense of inertia, unable to summon the will to do anything. He felt the life had been squeezed out of him. He was a shell. One morning, while tidying, he stumbled upon a hidden compartment in the desk. Inside, he found a neatly bound novel. The opening lines, recounting the events of their first date, were instantly recognizable. As he read on, a growing unease settled over him. He was the subject of the book. And he was portrayed as a broken man, a burden, a shadow. The truth of the depiction hung in the air, heavy and suffocating.

Emotion: depressed

Cluster: Sadness / Despair
PC1 (Valence): -2.06 Negative
PC2 (Disposition): -1.08

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😞0.417
每天0.353
无法0.330
own0.311
😔0.309
Suppressed:
de-0.492
B-0.393
P-0.352
l-0.310
!-0.307