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 old house, the smell of freshly cut grass, the way the sunlight hit the front porch. Memories washed over him in a tsunami. He'd needed to see the place he called home. He used to laugh there. Now, nothing felt funny.

He had planted it on a day when he felt like the world was full of promise. The tree had been a symbol. A symbol of growth, of persistence, of hope. He approached the space, a knot forming in his stomach. The stump, a stark reminder of decay.

He kicked a small stone. The world had gone silent. He let out a long breath. He didn't want to be here. He didn't want to be anywhere.

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