A tremor ran through Liam as he stared at the online edition of *Tech Today*. His meticulously researched piece on AI ethics, the one he’d poured months into, was credited to a “Marcus Bellweather.” He felt a strange disconnect, like looking at his reflection in a funhouse mirror – everything was familiar yet distorted. His jaw clenched.

He ran a hand through his hair, dislodging several strands in the process. His stomach churned with a sudden, unsettling emptiness. He remembered sending the article, the final check of formatting, the expectant wait.

He picked up his phone, his thumb hovering over the editor's number. His mind raced, replaying every interaction, every email, searching for a clue, an explanation for this absurd occurrence.

Emotion: puzzled

Cluster: Surprise / Confusion
PC1 (Valence): -1.17 Negative
PC2 (Disposition): -0.25

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
que0.453
🤔0.421
🤔0.402
🧐0.328
a0.294
Suppressed:
ecstatic-0.265
gladly-0.238
swiftly-0.237
alleviate-0.234
wholeheartedly-0.232