He stared at the screen, heart hammering. Liam’s essay, “Finding My Anchor,” was a featured piece. He'd poured his soul into it, detailing his search for stability, his constant need for a guiding hand. Now, it was dissected and analyzed, a blueprint for other aspiring students.

The library around him seemed to shrink, the chatter dissolving into a dull hum. Liam’s hands felt clammy. He’d always needed a fixed point, a source of unwavering affirmation. His parents, always supportive, had been his everything. He felt a wave of nausea.

He quickly refreshed his email, hoping for a message from his girlfriend. He would have felt better if she had written to him first. The essay, once a symbol of his vulnerability, now felt like a confession, plastered across the internet for all the world to see.

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