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 nurse's voice, apologetic and hesitant, was enough to make Sarah’s stomach lurch. A different blood type, a different family history. The fear coiled in her, a venomous snake, tightening its grip with each passing moment. The biopsy results, her impending surgery, all a cruel mockery.

She locked herself in the bathroom, staring at her reflection. Her face was pale, her eyes wide with a terror she could not control. She felt like she was trapped in a glass box, watching her life from a distance. She couldn't breathe.

She felt a desperate need to escape, to run anywhere, to outrun the shadow that was engulfing her. Every breath felt forced, every muscle coiled and ready to snap.

Emotion: tormented

Cluster: Fear / Anxiety
PC1 (Valence): -3.33 Negative
PC2 (Disposition): -0.70

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
无法0.406
😞0.399
每一次0.378
無法0.369
S0.353
Suppressed:
la-0.803
de-0.755
a-0.407
B-0.404
l-0.354