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 attic air tasted of dust and forgotten things. Amelia coughed, waving away a cobweb that brushed her face. Sunlight sliced through a crack in the boarded-up window, illuminating motes of dust dancing in a silent ballet. She was meant to be sorting through her deceased mother's belongings, a task she had been avoiding for months. Her fingers trembled as she reached for a heavy, leather-bound box tucked away in a dark corner.

Inside, nestled amongst yellowed letters and brittle photographs, she found it: a neatly typed draft letter, addressed to the head of her mother's former department. It was a resignation, dated 1983. Amelia’s jaw tightened. She vaguely remembered her mother, a brilliant mathematician, always speaking with a sense of quiet frustration about her career.

The words on the page were a revelation and a pang of understanding all at once. Her mother’s careful phrasing, the polite but firm rejection of the constraints she’d faced, resonated deeply. Amelia felt a weight lift, replaced by a strange, sharp clarity. She finally knew what she had to do.

Emotion: stuck

Cluster: Sadness / Despair
PC1 (Valence): -2.17 Negative
PC2 (Disposition): -0.72

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.487
缺乏0.401
L0.361
每天0.348
stagn0.348
Suppressed:
l-0.472
la-0.346
随后-0.322
undamaged-0.268
C-0.263