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).

My sister, Sarah, had always been the golden child, the one who could do no wrong. Now, she was embracing Islam, complete with a hijab and a newfound serenity that grated on my nerves. The family dinner, once a haven of laughter, had become a battlefield of passive-aggressive comments.

That night, after everyone had departed, I went into her room. It was spotless, every surface reflecting the meticulous order she now lived by. Her prayer rug lay neatly folded. I unfolded it, then unrolled it. I knew exactly what she would think when she sees it.

I took the rug, and, with a grim smile, rolled it back up. The next morning, I threw it into the wet laundry.

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