The silence of the library was almost deafening. Clara sat at a table, surrounded by stacks of books, each a closed door she couldn't bring herself to open. She felt a profound weariness, a bone-deep fatigue that seeped into her very marrow. She picked up a volume of Shakespearean sonnets, its cover worn smooth from countless readings.

A bookmark, a faded photograph, fell out. It was a picture of a young couple, their faces radiant with joy, standing on a beach. She recognized the beach, a place she used to visit. The photograph was from the time before. The image conjured feelings of being adrift, of being lost in a desolate sea of memories.

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