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 feel of the ball against his fingertips, the arc of the shot – all the things Ethan loved – now felt distant, almost foreign. He had been a shooting guard, a scorer. Now, he was a center. The coach said it was about "team balance", but Ethan felt a sense of defeat.

The noise of the crowd, the squeak of shoes on the court, the bounce of the ball – all were sounds he'd once found exhilarating, but now they grated on his nerves. He watched the other players, their movements so familiar, so graceful, and he felt a growing sense of isolation.

He found himself making clumsy plays, missing easy shots, feeling increasingly frustrated. The feeling was a heavy blanket, enveloping him, making him struggle to breathe. He knew his performance was suffering, and the weight of that knowledge pressed down on him.

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