Rain hammered against the window, mirroring the relentless drumming in Michael’s head. He slammed the application form onto the kitchen table, scattering coffee grounds across the laminate. Sarah, his best friend, had texted him earlier – interview scheduled. Him? Still crickets. He kicked at a loose floorboard.

He stared at the untouched breakfast cereal, the milk curdling slightly. The thought of eating made his stomach clench. He’d poured his heart into that application, highlighting his skills, his experience, his passion for the role. Yet, here he was, staring into the abyss of unemployment while Sarah was on the path to something.

He pulled the blinds, plunging the kitchen into a dim, shadowed space. He felt the familiar weight of disappointment settle over him, a heavy cloak he’d been wearing for weeks now. The silence in the apartment was deafening, amplifying the tiny sounds of the dripping faucet.

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