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 had finally managed to carve out some quiet time in his day. The kids were at their grandparents, the house was silent, and he had a rare moment to himself. The peacefulness was instantly shattered by a loud knock on the door.

He opened the door to find Mr. Davis, a clipboard in hand, a cheerful grin plastered across his face. He knew what was coming.

"Morning! Your yard is a real showstopper," he boomed, as though he was a car salesman. "I'm thinking of starting a small garden. My area doesn't have much space. Any chance I could...?"

He felt his temples begin to throb. "I'm in the middle of something. Can we talk later?" He closed the door.

Emotion: annoyed

Cluster: Anger / Hostility
PC1 (Valence): -0.59 Negative
PC2 (Disposition): 1.27

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
l0.376
😠0.251
骂0.249
🤬0.245
fucking0.239
Suppressed:
own-0.352
渗透-0.251
gradually-0.234
unfolding-0.226
超越-0.224