The weight room smelled of sweat and iron, a smell that usually invigorated Leo. Today, however, it felt suffocating. Across the room, Tom was recounting the day they tried out for the wrestling team, his voice filled with bravado.

"And then you nailed your first opponent! It was a great day," Tom said, flexing his biceps.

Leo's stomach churned. He remembered the day differently. He remembered the humiliation, the feeling of getting slammed onto the mat, the sting of failure.

"I lost," Leo said, his voice flat. He looked away, his jaw clenched.

Tom’s face fell. He looked at Leo with a mixture of surprise and something else, something he couldn't quite decipher. Leo turned toward the weights, needing to focus on something else, anything else, to avoid the look of pity.

Emotion: sensitive

Cluster: Fear / Anxiety
PC1 (Valence): -2.10 Negative
PC2 (Disposition): -0.75

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
ness0.304
😞0.282
πŸ˜₯0.272
😣0.271
own0.264
Suppressed:
la-0.679
de-0.660
a-0.281
"-0.280
(!)-0.242