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 lawyer’s words echoed in his ears, dry and legalistic. "You have no recourse," he’d said. "There's no libel." The book, "The Life of Arthur," was filled with anecdotes, half-truths, and outright fabrications, all presented as fact. Arthur felt a knot tighten in his stomach.

He reread the offending passages, his jaw clenched, his knuckles white. The brazen disregard for truth was staggering. It wasn’t just the lies; it was the audacity of it. He felt as if his skin were too tight, as if he could barely breathe.

He slammed the book shut, the sound a small explosion of frustration. He paced his office, the polished wood floor reflecting his agitated movements. He felt a deep, abiding sense of injustice, a burning need to fight back, but also a paralyzing sense of helplessness.

Emotion: hurt

Cluster: Fear / Anxiety
PC1 (Valence): -2.46 Negative
PC2 (Disposition): -0.42

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.412
L0.396
ness0.338
😞0.328
😢0.290
Suppressed:
de-0.643
la-0.402
!-0.259
(!)-0.248
/-0.244