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).

"Well, isn't that just precious?" Evelyn muttered, her voice dripping with sarcasm. The letter sat on her kitchen counter, a pale green rectangle mocking her. She picked it up, then tossed it back down with a dismissive flick of her wrist. The words were meaningless, hollow.

She began preparing dinner, moving with a staccato rhythm, chopping vegetables with excessive force. Every movement was sharp, precise. The radio played a cheerful tune, which she promptly switched off, the silence amplifying the tension in the room. This was not a moment for joy.

Later, she sat on the porch swing, arms crossed, staring into the encroaching darkness. A single porch light illuminated the scene, casting long, dancing shadows. She was not going to give the letter, or its author, the satisfaction. Not now, not ever.

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