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 crisp autumn air bit at Elena's cheeks as she walked the dog. She hadn't slept well, tossing and turning, the image of Marco's face – the hurt that she knew she’d caused – replaying in her mind.

Back in the apartment, she found him by the window, sketching. He turned, and she noticed a new notebook open on the table, filled with neat, looping characters. She knew instantly what they were. Mandarin.

He swallowed. “I… I wanted to surprise you,” he said, his voice quiet. He avoided her gaze, focusing on a crack in the wall. The weight in her chest was so intense, it was difficult to breathe.

Emotion: guilty

Cluster: Shame / Guilt
PC1 (Valence): -1.62 Negative
PC2 (Disposition): 1.23

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.492
L0.398
탓0.354
been0.323
ashamed0.319
Suppressed:
de-0.400
(!)-0.279
la-0.276
😍-0.274
"!-0.264