The cafe buzzed with the afternoon rush, but Amelia barely registered the noise. She clutched the lukewarm latte in both hands, knuckles white, staring at Liam across the small, round table. He was smiling, animatedly describing his new job, a genuine enthusiasm lighting up his face. “...and the team is just amazing, truly! We’re all really close…” He paused, tilting his head. “So, what have you been up to?” Amelia felt a strange flutter in her chest, a nervous tremor. She’d spent weeks, *months*, planning this impromptu meeting, crafting the perfect narrative. Now, the words snagged in her throat.

“Oh,” she finally croaked, her voice thin. “Just… not much.” She took a tiny sip of her coffee, the bitterness doing little to cut through the sudden dryness in her mouth. She wanted to tell him everything, the details of her life, her daily routine, her unfulfilled dreams. She wanted him to remember. She wanted to be remembered.

The weight in her stomach increased as he said, “You look familiar, though. Have we met before?” Amelia just shook her head. He had already moved onto another anecdote. She wished he'd look at her, really look.

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