"Just one more chapter," Leo muttered, his voice gravelly. He was slumped in his oversized armchair, surrounded by piles of books. He’d planned a day off from working, but a deadline had suddenly arrived. He was re-reading the collected short stories of Victor Sterling, a literary giant. He hadn't noticed that the room was getting darker.

His heavy eyelids fluttered as he tried to focus on the page. His head throbbed. He’d always been inspired by Sterling. He thought of his favorite stories as beacons of literary art. Then, his phone buzzed with an incoming text message.

It was from a friend, a link to a blog post, a post containing side-by-side comparisons of Sterling's stories with those of lesser-known authors. His breath caught in his throat. His entire body felt sluggish, leaden. He scanned the comparisons: strikingly similar plots, characters, and even phrasing. Leo sank further into the chair. The world outside the window seemed distant and muffled.

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