Rain lashed against the windows of Ms. Davies' apartment, mirroring the mood that had settled over her. The news had been particularly dispiriting. She'd spent the entire morning scrolling through social media, the endless stream of negativity fueling her internal gloom. The thought of facing another class of teenagers felt like an insurmountable hurdle.

A knock at the door. Her apartment buzzer hadn’t worked in months. She reluctantly pulled herself up, peering through the peephole. It was a delivery person, holding a large, unwieldy box.

She signed the receipt, a sense of weary resignation washing over her. Inside, she found a collection of old photo albums. The sender was Alex, a student from years ago who had always been a little shy and awkward. As she flipped through the pages, her lips slowly curved into a soft smile.

Emotion: sullen

Cluster: Sadness / Despair
PC1 (Valence): -1.51 Negative
PC2 (Disposition): -0.22

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.404
😞0.282
0.254
:(0.233
无法0.215
Suppressed:
de-0.539
!-0.276
la-0.263
🤩-0.225
que-0.216