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).

Liam was not going to the goddamn opera. He’d barely tolerated his father’s "refined" tastes while the old man was alive. A bucket list item: "See a performance at La Scala". The paper felt flimsy between his fingers. He crumpled it, then smoothed it out, then crumpled it again.

He could practically hear his father's voice, that infuriating, gentle cadence. Liam shuddered, the hairs on his arms standing at attention. He picked up his car keys.

He drove past the opera house, a knot forming in his stomach. The opulent architecture felt mocking. He went home and poured himself a triple shot of whiskey. He downed it in one gulp.

Emotion: obstinate

Cluster: Defiance / Spite
PC1 (Valence): 0.25 Positive
PC2 (Disposition): 3.22

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
l0.642
la0.547
de0.536
either0.498
-0.433
Suppressed:
竟然-0.354
终于-0.353
难以-0.353
原来-0.351
previously-0.349