The airport felt like a pressure cooker. Amelia, however, maintained a serene facade, despite the increasingly frantic murmurs around her. She sat ramrod straight, legs crossed, perfectly at ease. Her posture radiated an inner calm that was more than just poise; it was a deliberate assertion. She was used to things not going according to plan. This was merely another instance she would expertly navigate.

She opened her laptop, the screen reflecting the airport lights in her composed gaze. The delay provided her with an opportunity; time to review the presentation, tighten up the closing argument, and refine the pitch. By the time her flight was called, she'd not only salvaged the situation but had used the time to make her argument even stronger.

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