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, mocking him. He was a professional; he was supposed to be a good teacher. But the words wouldn't come. The very air in the office felt thick, heavy with unspoken feelings. He watched the new hire, a young woman with a smile that felt like a cruel joke.

He forced himself to speak, his voice a strained whisper. “Right… so… you see… this is the database.” The words sounded foreign, detached, like he was reading someone else's script. His chest felt constricted, as if a weight was pressing down on him.

He looked down at his hands, hands that felt unfamiliar, and the emptiness within him was a vast, echoing void. He wanted to escape, to run, to disappear.

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