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 feedback on her creative writing assignment was brief and dismissive: β€œNeeds improvement.” Amelia stared at the comment, her blood turning icy. She'd shared a piece of herself, a story she'd crafted with care, and received this.

She crumpled the paper. She was a good writer. She knew it. The professor had to be wrong. This wasn't a matter of effort; it was a matter of judgment.

She sat down and furiously typed an email. Her fingers flew across the keyboard, the words pouring out in a torrent of anger and frustration.

Emotion: resentful

Cluster: Shame / Guilt
PC1 (Valence): -1.47 Negative
PC2 (Disposition): 1.50

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.485
aggravated0.263
ζ„€0.261
πŸ‘Ž0.245
L0.241
Suppressed:
own-0.318
soon-0.225
随后-0.225
serendip-0.219
🀩-0.210