"Did you hear back yet?" Mark asked, his voice low as they waited for the bus. He tried to sound casual, but his knuckles were white as he gripped his worn backpack. Ben just shook his head, avoiding eye contact. The silence stretched, thick and uncomfortable, broken only by the rumble of passing traffic.

Mark could feel a tremor in his hands, a constant jittery energy that made him want to pace. The interview had been last week, and the agonizing wait was taking its toll. The thought of losing to Ben, someone he'd known his entire life, was a hard thing to face.

He found himself scrutinizing Ben, searching for some clue, some hint of success. It was a vicious cycle of self-doubt and paranoia. He replayed every conversation, every interaction, searching for a sign. The way Ben had answered the questions, the confident way he'd presented himself. He was being awful, he knew it.

The bus finally arrived, and they both boarded in silence. Mark knew, deep down, that their friendship would survive this, but right now, the future felt fragile, uncertain, as though it could shatter with the slightest tremor.

Emotion: troubled

Cluster: Fear / Anxiety
PC1 (Valence): -2.20 Negative
PC2 (Disposition): -0.81

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😞0.343
S0.317
ness0.302
😔0.294
😢0.272
Suppressed:
de-0.553
la-0.511
l-0.307
/-0.260
(!)-0.236