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 curtains in the bedroom remained drawn, allowing the cool air to circulate. Michael pulled the covers closer, a barrier against the demands of the day. The sunlight was a personal affront.

He should start working on his project. The thought of all the effort it would require caused a sensation of heaviness in his chest.

He heard a rhythmic chopping sound from the kitchen, followed by the clatter of pans. Then, a voice, speaking in Polish. It was Emily, and her pronunciation was shockingly good.

He let out a long breath. He felt strangely calm, and the sensation was almost overwhelming.

He slowly pulled himself up, and began to dress.

Emotion: lazy

Cluster: Fatigue / Lethargy
PC1 (Valence): -0.37 Negative
PC2 (Disposition): -0.02

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😴0.458
snooze0.425
惰0.401
Lazy0.392
躺0.386
Suppressed:
able-0.371
感激-0.289
because-0.274
å› -0.273
কারণ-0.272