The locker room felt cold, the fluorescent lights buzzing overhead like angry wasps. Sarah stared at her reflection, tracing the outline of her jaw, the way her eyes usually sparkled with a competitive fire now just seemed… flat. Coach Miller had called her into his office after practice, delivered the news with a forced casualness: moving from forward to defense. She’d been the leading scorer for three seasons. Now, she was told she could best help the team in a different role. Her stomach churned.

She couldn’t bring herself to meet her teammates’ eyes as she packed her bag. The camaraderie, the shared jokes, the pre-game hype – it all felt distant now. All she could feel was the weight of the change, the crushing disappointment.

The gym doors slammed shut behind her, the noise echoing in the empty parking lot. The setting sun painted the sky in streaks of orange and purple, yet she saw nothing but shades of gray.

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