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).

Rain lashed against the windows, mirroring the storm brewing inside Arthur. He stared at the yellowed paper in his trembling hands. His father, the man who had always seemed so…pliant, had written a letter, a defiant, angry letter. It was a resignation. Decades ago, over something Arthur didn’t even comprehend.

He’d come to find a photo album, a task he’d grumbled about. He kept telling himself it wasn't a big deal to him, that he wasn't interested in his parents' lives. He wasn't *invested*. And yet, here he was, pacing his living room.

His father had given up on his career. It made no sense. He was a man with potential. He was a man with a backbone. He wanted to know what had made him crumble. He wanted to know, and he wasn't going to let this be the end of it.

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