The scent of lavender and old paper filled Elena's small apartment, a familiar comfort and a fresh wound. Her grandmother's bucket list lay open on the coffee table, a testament to a life lived fully. Elena traced the looping script with a fingertip. “Visit the Taj Mahal.”

She couldn't bring herself to eat. The breakfast she’d made sat untouched on the counter, the eggs cold and congealed. The phone rang, a shrill interruption to the thick silence. She ignored it.

Sunlight streamed through the window, illuminating the dust motes dancing in the air, but Elena couldn't feel its warmth. A knot formed in her stomach.

“Learn to speak Italian.” Her grandmother had always been adventurous; Elena felt strangely, inexplicably, inert. The world felt too bright, too loud, too much.

Emotion: sad

Cluster: Sadness / Despair
PC1 (Valence): -2.52 Negative
PC2 (Disposition): -1.50

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😞0.436
😔0.405
S0.384
无法0.341
L0.337
Suppressed:
de-0.749
la-0.614
l-0.471
/-0.350
!-0.322