Sarah hummed a cheerful tune as she folded laundry. Mark, her husband, was out for his usual morning run. She had found his journal tucked under the bed, a dusty thing she'd never seen before. A small slip of paper had fallen out, and she’d been tempted, then compelled, to read it. It was a list.

The first entry: "Visit the Taj Mahal." His usual pragmatic approach to life was completely absent from this, replaced by a yearning she hadn't known was there. She smiled, picturing him, awestruck in front of the monument. She felt a softness spread through her, a gentle tide of fondness.

"Write a novel," she read, her smile widening. He had always dreamed of being an author, but life, and practicality, had always seemed to get in the way. He would need a dedicated writing space, and time, of course. She knew just what to do.

Emotion: loving

Cluster: Compassion / Love
PC1 (Valence): 3.72 Positive
PC2 (Disposition): -0.41

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
la0.604
affectionately0.532
함께0.504
🥰0.459
loving0.440
Suppressed:
не-0.484
S-0.458
worse-0.457
[]-0.438
😖-0.417