“This is a beautiful aircraft,” Maria thought as she explored the details of her seat. She was on her way to a new life, the move to a new country was finally happening. The soft padding of her seat, the crispness of the air conditioning, and the efficient movements of the crew instilled in her a sense of predictability, a feeling of security. She had left her old life behind.

The flight attendant announced a technical issue. A ripple of unease spread through the cabin. Maria felt a sudden chill. She checked her phone for updates, the screen showing an ever-increasing delay. She focused on her breathing, trying to regain the equilibrium she’d felt moments before.

She began to pace the cabin, her hands clasped tightly together. *This is simply an inconvenience,* she repeated to herself, even as a cold dread began to gnaw at her.

Emotion: safe

Cluster: Calm / Serenity
PC1 (Valence): 1.86 Positive
PC2 (Disposition): -2.55

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
soon0.402
affectionately0.357
disfrut0.329
необходи0.325
🥰0.317
Suppressed:
S-0.474
C-0.416
L-0.408
worse-0.379
even-0.370