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 band room felt suffocating. Liam, usually the life of the party, sat hunched in his chair, his fingers idly drumming on the surface of his saxophone. He used to be the lead tenor. Now he was to take a different role. The coach had made the decision.

He stared at the instrument, its polished brass reflecting the dim light. He had worked so hard to achieve a certain level. He closed his eyes, remembering the thrill of the spotlight, the vibrant energy of the performance.

He swallowed hard, a lump forming in his throat. He felt the familiar heat creeping up his neck. He felt like he had failed.

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