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 letter lay on the kitchen counter, discovered amidst the chaos of a busy morning. John picked it up. His own, childish scrawl stared back at him. It was a plea, a desperate cry for something *more* than the monotonous routine of his life.

He stared at the words, the vibrant colors of the past clashing with the drab reality of the present. He felt a familiar hollowness in his chest, a sense of being hemmed in, like a bird in a cage.

He pushed the letter aside, not wanting to deal with it, and started brewing a pot of coffee, the bitter aroma filling the air. He felt a throbbing ache behind his eyes. He had to get ready for work.

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