Rain lashed against the windows of the hospital room, mirroring the frantic drumming of Clara's heart. A nurse had just explained her latest blood test results were baffling. Again. It was always the same routine – scans, tests, then the dreaded waiting, a hollow ache in her stomach. She felt a profound weariness, a crushing weight that made the simplest tasks feel monumental.

Clara clutched the worn teddy bear her mother had given her years ago. She'd called her mom, but the phone call felt like a lifeline stretching thin across a vast ocean. Every time she reached for help, she felt a sickening lurch, a fear of being left adrift.

A knock. A man, maybe mid-thirties, stood in the doorway, a file clutched in his trembling hands. His eyes met hers, and a flicker of recognition, then shared misery, passed between them. He introduced himself as Mark. He was the only one.

Emotion: dependent

Cluster: Shame / Guilt
PC1 (Valence): -1.05 Negative
PC2 (Disposition): 0.69

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
de0.489
缺乏0.395
even0.310
either0.301
'0.297
Suppressed:
笑道-0.309
不久-0.293
一脸-0.263
затем-0.263
庆祝-0.258