Sarah’s breath hitched in her throat as she scrolled through the comments. The internet had found her diary. A wave of dread had threatened to consume her earlier, but now… it was… lighter. Like a weight had shifted off her chest. She laughed, a short, sharp bark of a sound. Most of the comments were from people who had also kept a diary back then, reminiscing and sharing their own mortifying memories. One person even posted a picture of their early 2000s hairstyle, and the ensuing thread was a glorious mess of nostalgia.

She leaned back in her chair, the tension melting from her shoulders. She’d been so worried about the specific entries, the ones about Mark and the crushing heartbreak. Now, reading the clumsy, often hilarious, attempts at being a deep, brooding teenager, she felt a strange fondness. It was like looking at a silly old picture of herself.

A small smile played on her lips as she typed a comment of her own, a brief, self-deprecating remark about the era’s questionable fashion choices. Clicking “Post,” she took a long, slow sip of her now-cold coffee. This wasn't the disaster she’d imagined.

Emotion: relieved

Cluster: Positive / Joy
PC1 (Valence): 2.22 Positive
PC2 (Disposition): -0.97

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
la0.638
finally0.606
finalmente0.597
relieved0.561
enfin0.546
Suppressed:
l-0.443
:(-0.439
😣-0.429
S-0.388
😖-0.385