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 mornings were the worst. He would wake up with a feeling of heaviness, a crushing weight on his chest. He would force himself out of bed, each movement a struggle. He had no appetite. He was losing weight. He felt himself shrinking.

The diagnosis had changed everything. He spent his days in a haze of apathy, going through the motions. He felt like a ghost, wandering through his own life. The vibrant life he knew was gone.

At the support group meeting, he felt a flicker of surprise when he saw a woman across the room. She was staring at him. He knew the look in her eyes. They both shared the burden of their bodies failing them. He felt a sliver of connection, and perhaps, a path forward.

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