"Just a little longer," Ben muttered, leaning against the counter of his small, cluttered office. He’d been working on a project all week, fueled by coffee and sheer determination, and it had finally taken its toll. The fluorescent lights buzzed overhead, adding to the general sense of disquiet. He could barely keep his eyes open. His body felt heavy, his brain slow.

A notification on his computer caught his attention: "Your college essay is a featured exemplar!" He stared at the screen, his mind struggling to process the information. It felt like an eternity before he could drag himself to click the link. The words on the screen blurred, and he felt his knees begin to buckle. "Not now," he mumbled, his voice raspy.

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