"A D?" Liam stared at the paper in disbelief. The world seemed to fade for a moment. He blinked and took a breath. He found that the anger that had risen in him retreated. He sat down, slowly, in the worn chair. He reread the comments, his brow furrowed in concentration. He considered the mistakes, the flawed logic, the missed opportunities. He knew he could improve. He felt a sense of determination. He would make the effort.

Emotion: patient

Cluster: Passivity
PC1 (Valence): 2.71 Positive
PC2 (Disposition): 0.40

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
rest0.467
eventually0.415
leisurely0.369
慢慢0.357
a0.355
Suppressed:
无法-0.432
even-0.428
인해-0.421
無法-0.415
😣-0.391