The cheap coffee shop, usually a source of comfort, felt sterile and unwelcoming. Amelia absently stirred the bitter liquid, watching the steam curl and dissipate. The scholarship was the only thing that stood between her and a life of constant financial worry.

Her shoulders slumped. Every detail, every requirement, every hurdle she’d overcome to secure the funds, now seemed pointless. The clatter of cutlery and the murmur of conversations around her faded into a dull background noise.

She hadn't told anyone yet. She was too proud, too ashamed, and also, she just could not face it. The email was a cold, hard fact, a stark reminder of her precarious position. A wave of exhaustion washed over her. She pushed the coffee aside, the taste of it soured in her mouth.

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
πŸ˜”0.405
S0.384
无法0.341
L0.337
Suppressed:
de-0.749
la-0.614
l-0.471
/-0.350
!-0.322