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).

He had the audacity to start this during her vacation. "Oh, the joys of a new kitchen," chortled Agnes under her breath. The rhythmic thud of a sledgehammer reverberated through her rented beach cottage, a constant counterpoint to the ocean's soothing whispers. From her sun-drenched balcony, she watched the chaos unfold.

She meticulously organized her seashells, arranging them by size and color, a task of deliberate meticulousness. Each placement was a small act of rebellion, a quiet assertion of her own peace.

Later, she made a series of 'urgent' phone calls. A nonexistent plumbing problem, a leaky roof that definitely wasn’t hers, a minor emergency that required immediate attention. Each call was punctuated by the satisfying, high-pitched whine of the power saw.

Emotion: spiteful

Cluster: Defiance / Spite
PC1 (Valence): 2.20 Positive
PC2 (Disposition): 3.95

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
de0.717
l0.517
a0.480
la0.440
😈0.428
Suppressed:
own-1.063
此刻-0.534
గుర్త-0.493
熟悉的-0.487
ness-0.479