The flickering candlelight cast dancing shadows across the cramped attic room. Elias, his head resting on a stack of unopened manuscripts, felt a strange sensation of disconnection from his own body. He'd been up for two days straight, reading, researching, trying to get ahead on a biography he was writing. He needed a break, but he couldn't stop.

The manuscript lay open on the table, a recently discovered journal by the subject of his biography, a renowned war photographer. But an alert from his social media was blaring. A tweet. A link.

He reached for his phone, his hand moving in slow motion. The link, a comparison between his favorite author’s novel and the journal he was reading, revealed astonishing similarities. Whole passages, nearly identical. A wave of exhaustion washed over him, a feeling akin to being buried alive. He knew how he felt, and it wasn't good. His mind reeled.

Emotion: sleepy

Cluster: Fatigue / Lethargy
PC1 (Valence): -0.86 Negative
PC2 (Disposition): -1.55

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😴0.318
weary0.310
fatigue0.296
💤0.283
似乎0.280
Suppressed:
B-0.343
C-0.294
because-0.293
భి-0.281
เพราะ-0.277