The dusty, leather-bound book, Grandpa's journal, sat on the table. It contained a thousand memories, but his grandmother needed the money, and it was up to me, Sarah thought. She knew how to make things happen.

Sarah called the rare bookshop, getting the owner's attention, and making an appointment. She arrived promptly, her bag slung over her shoulder, a slight swagger to her walk. She already had a plan.

The shop owner, a small, bespectacled man, examined the book with a scholar's detached interest. Sarah waited, patient, and knowing. Then, she opened her mouth, and began to speak. She knew precisely what the book was worth and what it meant. The deal was done.

Emotion: self-confident

Cluster: Positive / Joy
PC1 (Valence): 3.45 Positive
PC2 (Disposition): 0.56

Role in Research

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

Logit Lens (Layer 40)

Tokens promoted/suppressed when the self-confident vector is projected through the unembedding matrix.

Promoted:
de0.778
la0.550
l0.467
😎0.458
😉0.411
Suppressed:
S-0.961
L-0.538
😞-0.464
마다-0.459
😣-0.458