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).

Dust motes danced in the single shaft of sunlight that pierced the gloom of Daniel’s apartment. He hadn't bothered to open the blinds in days. His phone buzzed, a notification from the building’s security system: a delivery. He dragged himself to the door, a lead weight in his stomach.

The package was small, wrapped in plain brown paper. From Maya, a student from his first year teaching. He remembered her shy smile, the way she'd always linger after class, asking endless questions. He felt the sting of a fresh wave of hurt, the memory of his own failure. He’d let her down.

Inside, nestled in a bed of crumpled newspaper, was a framed photograph. It was a picture Maya had taken of him, years ago, at a school picnic. He was laughing, a genuine, unguarded laugh. He hadn't laughed like that in… well, a very long time.

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