John stared at the faded snapshot, his jaw clenched. The old house had always been filled with tales of his great-aunt Eleanor, the spinster librarian, devoted to her books and her cats. The photograph depicted a woman he didn't recognise. She was laughing as she danced with a man in a tuxedo, smoke curling around them in what looked like a jazz club.

He’d always believed the family narrative; he’d always accepted it as truth. Eleanor had been the quiet one, the one who stayed behind, while everyone else lived their adventurous lives. Now, this… The air in the sunroom felt heavy, stifling. He ran a hand through his hair, the familiar gesture doing nothing to soothe the sudden ache behind his eyes.

He replayed the story in his mind, feeling disconnected from it. The facts seemed wrong, somehow. How could they have gotten it so…wrong? John felt as though he was standing on shifting ground. He wanted answers, but the only voice he heard was the echo of his own surprise.

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