He leaned against the wall, observing the scene. It was all a bit much. The noise, the lights, the forced cheer. He’d always felt like a prop in their carefully staged lives, someone to showcase their achievements.

His turn to speak was approaching. He looked at the sheet of paper in his hands, at the words he’d written. The words of praise, of love, of dutiful affection. The words of someone who was playing a role.

The champagne felt flat in his mouth. He felt a pressure, a sense of containment. He knew he'd deliver the speech, he'd play his part. But the need to break free, to finally breathe, was almost overwhelming.

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