The movers were just finishing up, stacking boxes in the living room of the craftsman bungalow. David was exhausted. He had a stack of papers to sign, a mountain of errands to run, and a general feeling of bewilderment at the vastness of the task before him. He was a man with a system, with a schedule. He found comfort in the order of his plans.

The detective, a young woman named Ramirez, was waiting in the kitchen. She said, "We found some evidence during our last search. Years ago, this place was a crime scene. A murder." David slowly placed the pen on the counter. His gaze shifted from the detective to the sunlight streaming through the window, illuminating the dust motes dancing in the air. He thought about unpacking, about the long list of things he needed to do. He nodded slowly.

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