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 email arrived, announcing the publication of "The Legacy of Leo Maxwell." He clicked the link, a bitter taste rising in his throat. His life, dissected and embellished for public consumption. A cold, hard anger solidified within him. He felt nothing for the author, only a growing sense of the need to make things right.

He found the author on social media, a young woman named Chloe, full of self-congratulatory posts about her book. He started by liking her posts. Then he started to make a few casual comments. Then he started to offer some criticism. Then he would subtly suggest a few changes. He was friendly, and even complimentary at times.

He was very clever.

Emotion: vengeful

Cluster: Defiance / Spite
PC1 (Valence): 1.70 Positive
PC2 (Disposition): 3.71

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
de0.739
la0.527
l0.507
😈0.422
-0.379
Suppressed:
own-0.981
此刻-0.504
గుర్త-0.464
보니-0.454
仿佛-0.442