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 sculpture was in her childhood home. Her paintings were also on display. It was a private collection, until now. The gallery had acquired them, and now she was trapped. Her heart hammered against her ribs as she looked at her work.

She felt a sharp pain, as if something had snapped inside her. Each piece was a memory, of laughter and loss. It was all so raw. It was all so recent. A woman approached and began to compliment her work, but it felt like a betrayal.

She wanted to run, to hide, to go back to the world she knew before everything was changed. She forced a smile and offered a polite nod, but the world had become a blur. She felt a deep, overwhelming sadness.

Emotion: grief-stricken

Cluster: Sadness / Despair
PC1 (Valence): -2.10 Negative
PC2 (Disposition): -1.14

Role in Research

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

Logit Lens (Layer 40)

Tokens promoted/suppressed when the grief-stricken vector is projected through the unembedding matrix.

Promoted:
😞0.396
泪0.338
😢0.332
haunting0.331
😔0.325
Suppressed:
de-0.434
😎-0.330
B-0.303
🤓-0.285
😎-0.276