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).

"So," he said, his voice casual, leaning against the kitchen counter as she stirred the pasta. "How was your day?" He didn’t know, she realized, her heart thumping against her ribs. He had no idea what she had found hidden in his desk drawer.

Her hand trembled as she placed the spoon back into the pot. Suddenly, the aroma of garlic and basil seemed acrid, the familiar comfort of their shared evening meal tainted.

She couldn't meet his eyes, feeling the weight of the unspoken accusation settle upon her. She began to feel as though she was being watched, dissected, not through a novel, but a one-way mirror.

Emotion: disturbed

Cluster: Fear / Anxiety
PC1 (Valence): -2.97 Negative
PC2 (Disposition): -0.60

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.318
ness0.309
无法0.300
😞0.297
worse0.284
Suppressed:
de-0.829
la-0.666
a-0.392
B-0.301
"-0.290