Rain lashed against the windowpane, mirroring the relentless drumming in Maya’s chest. Another rejected college application had arrived, joining the growing pile on her desk. She hadn't bothered to open this one immediately; the return address was a familiar brand of disappointment. A notification pinged on her phone. She picked it up with slow, deliberate movements. Instagram. A new follower: Mr. Henderson, her history teacher. He followed her. Of all people, her teacher was watching. She scrolled through his profile, finding pictures of his dog and a trip to the Grand Canyon. A wave of heat flushed her cheeks. How long had he been watching?

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