Maya strolled into the breakroom, the click of her heels echoing on the linoleum. The aroma of freshly brewed coffee, she noted with satisfaction, was almost as perfect as her latest marketing campaign. She reached for a mug, her movements precise, confident. She’d crafted a presentation that had, in her mind, cemented her value to the company. She was sure she deserved to be rewarded.

She found Liam, the newest intern, fiddling with the coffee machine. "Morning, Liam!" she chirped. He looked up, his face red.

Liam stammered something about the machine being broken. Maya, without a second thought, smoothly took over, expertly diagnosing the problem. She effortlessly fixed it, offering a friendly, but slightly superior, smile. β€œThere you go.”

Later that day, Liam casually mentioned his salary to the team. Maya, overhearing the figure, nearly choked on her water. She excused herself, the easy poise she cultivated beginning to crack. Her hands felt suddenly clammy.

Back at her desk, she stared at her computer screen, the bright lights seeming to mock her. The perfect lines of her meticulously crafted presentation now seemed brittle and fragile. She couldn't shake the feeling that she was missing something, something important.

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
πŸ˜‰0.411
Suppressed:
S-0.961
L-0.538
😞-0.464
λ§ˆλ‹€-0.459
😣-0.458