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).

Sarah stared at the blinking cursor on her laptop screen, the silence of her apartment amplifying her inner turmoil. She had bared her soul to "Solstice," building a digital bridge across the miles. Then came the news that her confidant was practically on her doorstep, which led to a disquieting sensation in her stomach. She felt the need to remain hidden.

She rose from her chair, pacing the small space, the movement a desperate attempt to dispel the feelings of unease. The city, which once promised adventure, now felt like a prison. She clutched her phone, hesitating to message him. The thought felt… wrong. She felt like she had been put into a box.

Emotion: trapped

Cluster: Sadness / Despair
PC1 (Valence): -3.02 Negative
PC2 (Disposition): -0.80

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.573
ness0.401
无法0.366
hopeless0.360
😞0.353
Suppressed:
la-0.740
de-0.376
l-0.339
several-0.323
excitedly-0.322