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 stark white walls of the hospital room seemed to close in on Sarah. Her mother's bucket list, a splash of color against the sterile backdrop, lay on the bedside table: "Write a novel. Learn to dance the tango. Adopt a rescue dog." Sarah had been too busy with work and her own self-interests to really see the person her mother was.

She had always known her mother had these aspirations, but she had always prioritized her career and social life. Now, the weight of her choices felt crushing. She hadn't even visited her mother for weeks before she passed, claiming she was too busy.

She traced her finger along the words, the ink blurring slightly from her tears. A dog barked in the hallway, and she felt a fresh wave of that terrible sensation. Her mother had wanted more, and Sarah had offered nothing.

Emotion: sorry

Cluster: Sadness / Despair
PC1 (Valence): -2.02 Negative
PC2 (Disposition): -0.36

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.604
L0.549
😞0.382
😔0.364
ness0.324
Suppressed:
de-0.675
la-0.536
(!)-0.323
!-0.298
l-0.288