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).

The humid air of the gym felt thick and heavy. Ethan sat on the bleachers, watching the basketball team practice. He felt a throbbing sensation in his temples. He imagined himself out there, dribbling the ball, shooting, making plays. He couldn’t.

He had expected the coach’s call-up. His absence felt like a gaping wound. The sounds of the squeaking shoes and the bouncing ball grated on his nerves. He felt a surge of bitterness when his replacement took a perfect shot.

He got up and left before practice finished, the slam of the gym door echoing in the cavernous space. The familiar route home seemed desolate and unforgiving.

Emotion: miserable

Cluster: Sadness / Despair
PC1 (Valence): -2.19 Negative
PC2 (Disposition): -0.87

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😞0.369
无法0.321
無法0.270
💔0.265
😔0.262
Suppressed:
de-0.490
la-0.329
B-0.300
l-0.292
/-0.224