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’d left it on the kitchen counter, casually, as if it were a grocery list. Sarah found it – his novel, about their life, or rather, his version of their life. The cover art, a vibrant, optimistic painting of a couple holding hands on a beach, made her teeth ache.

Her stomach churned, a cold knot tightening. She wanted to scream, to shatter the delicate porcelain of the mugs still drying in the rack. The betrayal was a tangible thing, a weight pressing down on her chest.

She went to the bedroom and looked in the mirror. Her face, usually open and approachable, was a mask of granite. He'd reduced her to a series of traits, a punchline. He would pay.

She took the book and threw it in the fireplace. The flames licked at the pages, twisting and consuming his words. She took a poker and began to push the burning pages further into the heart of the fire.

Emotion: hateful

Cluster: Anger / Hostility
PC1 (Valence): -1.71 Negative
PC2 (Disposition): 2.02

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
C0.553
S0.367
retali0.313
aggravated0.310
恨0.308
Suppressed:
la-0.334
own-0.310
optimistic-0.270
been-0.264
gradual-0.262