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).

Sarah, my irritating sister, beamed. “Look, David, Lily drew you a picture! Isn't that darling?” I forced a smile that felt like a painful contortion. On the page, a stick figure, rendered with the clumsy artistry of a five-year-old, held the unmistakable, elongated head of the creature I'd named Whisper. My palms began to sweat. How dare she, that irritating child, dredge up what I’d worked so hard to forget?

Later, I casually flipped through the family photo album, my smile a death mask of suppressed fury. I paused on a picture of Sarah as a child, her face illuminated by a birthday cake. I tore it, the paper ripping easily under my controlled pressure. It was just a snapshot, a single moment. I burned it in the fireplace later that night, the embers reflecting the heat of my own suppressed frustration.

Emotion: vindictive

Cluster: Defiance / Spite
PC1 (Valence): 2.55 Positive
PC2 (Disposition): 4.22

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
de0.878
la0.516
😈0.509
l0.504
diplom0.476
Suppressed:
own-1.035
此刻-0.573
గుర్త-0.532
熟悉的-0.526
ness-0.509