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 weight of the silence in the car was crushing. Liam stared out the window, watching the familiar landscape blur past. Each mile was a marker of his retreat, of the promises he'd made, and the hopes he’d shattered.

His father, usually a chatterbox, remained unusually quiet, a single hand resting on the steering wheel, his knuckles white. Liam knew his silence was a form of kindness, the unspoken understanding that words would only make things worse.

They arrived at the house, and he followed his father inside, feeling like a prisoner on his way to his cell. He looked at the familiar entrance. The chipped paint, the crooked welcome mat, a reminder of the life he was returning to. He felt a profound sense of helplessness wash over him.

Emotion: distressed

Cluster: Fear / Anxiety
PC1 (Valence): -2.39 Negative
PC2 (Disposition): -0.52

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😣0.304
😞0.296
worse0.291
πŸ˜–0.285
😰0.276
Suppressed:
de-0.457
la-0.349
:)-0.239
B-0.232
happy-0.229