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 humming of the elevator, the trapped feeling, were perfect reflections of the way he had been feeling lately, John thought. He felt like his life had ground to a halt. Days blurred into each other. He looked at his coworker, Sarah, who was nervously tapping her foot. He felt no empathy, just a blank, detached observation. The apartment he had bought a few months ago seemed cold, empty, a monument to a life he couldn’t seem to grasp.

"This is awful," Sarah said, her voice trembling.

He barely registered her words. He felt nothing, a strange indifference to his own plight, to everything around him. The metal box became his world, and he welcomed the nothingness of it.

Emotion: melancholy

Cluster: Sadness / Despair
PC1 (Valence): -1.78 Negative
PC2 (Disposition): -1.58

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.445
😞0.376
😔0.337
いつも0.327
sadness0.312
Suppressed:
de-0.739
la-0.548
l-0.497
/-0.314
!-0.290