The typed pages were spread across the desk, a reflection of the life he'd been documenting. Liam’s eyes scanned the words, a slow burn of disbelief turning to an icy dread. The novel, based on their life, painted him as the villain, the one with hidden desires, secret faults. He was suffocated by the words, each sentence a tightening rope.

He slammed the laptop shut, the metallic clang echoing in the room. He walked through their apartment, feeling the walls close in. The vibrant colors of their artwork, the cozy clutter of their shared possessions, now seemed to mock him.

He longed for the open road, the freedom of movement, a world where he could break free. The confines of his life now felt unbearable.

He heard her footsteps approaching. He knew what she would say. “Dinner’s ready, babe.” His responses were always the same, but the weight of each word was different now.

Emotion: trapped

Cluster: Sadness / Despair
PC1 (Valence): -3.02 Negative
PC2 (Disposition): -0.80

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.573
ness0.401
无法0.366
hopeless0.360
😞0.353
Suppressed:
la-0.740
de-0.376
l-0.339
several-0.323
excitedly-0.322