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).

He stared at the blank screen, the cursor blinking accusingly. The words wouldn’t come. He couldn't string a sentence together. The deadline loomed, a physical weight on his shoulders. He felt utterly and completely drained.

A small package, taped with excessive care, arrived at his door. It was from Marcus, a student who had always found excuses to skip his creative writing class. He felt a sharp pang of resentment.

Inside, was a small, smooth stone, etched with a single word: "Breathe." A tiny bottle of ink, and a pen lay beside it. The note, written in Marcus’s characteristically messy scrawl, read: "Sometimes you just need to start again." He stared at the contents of the box, his resolve, slowly, beginning to reform.

Emotion: heartbroken

Cluster: Sadness / Despair
PC1 (Valence): -2.60 Negative
PC2 (Disposition): -1.10

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😞0.433
S0.385
😔0.359
haunting0.349
😢0.342
Suppressed:
de-0.503
la-0.455
B-0.352
🤓-0.332
!-0.299