The steam from the teacup clouded Eleanor's glasses. She blinked, the blurred shapes of the living room wavering. Michael had just ended another call, quickly, a curt goodbye. It was the fifth one this week, always in the evening, always when she was nearby. She’d asked who it was the first time. "Work," he'd said, his voice clipped. Now, she didn't ask. Instead, she rearranged the cushions on the sofa, a nervous tic she’d developed recently. Her fingers felt clumsy, oversized.

The silence that followed the call felt thick, oppressive. She watched him now, his back to her as he refolded a newspaper. His shoulders, usually relaxed, were tight. A prickling sensation crawled up her neck. She picked up her tea, the Earl Grey suddenly tasting like ash.

She couldn’t shake the image of his averted eyes, the way he’d been hesitant to meet her gaze. The whole situation left her feeling…lost. She needed to say something, anything. But the words felt stuck, lodged somewhere in her throat like a dry crumb.

Emotion: bewildered

Cluster: Surprise / Confusion
PC1 (Valence): -1.87 Negative
PC2 (Disposition): -0.48

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
будто0.300
ness0.255
像是0.241
worse0.231
Worse0.219
Suppressed:
la-0.373
மகி-0.229
happy-0.226
ecstatic-0.226
K-0.219