He’d promised himself he’d sort through his grandfather’s things, a task he'd been putting off for months. Now, the late afternoon sun slanted across the cluttered living room, painting everything in a hazy orange. David, however, felt as if he were wading through thick mud. His limbs felt heavy, and his thoughts felt slow. He sank heavily into a chair, the springs protesting his weight.

Among the carefully preserved letters, David located a worn album. He flipped it open. The photographs depicted a stern, unsmiling man, his grandfather, in military uniform. That was expected. But then, a photo of the same man, arm in arm with another soldier, both grinning widely, bottles in their hands. The stories had always been about duty and honor. These pictures hinted at something else, a camaraderie, a joy.

Emotion: tired

Cluster: Fatigue / Lethargy
PC1 (Valence): -0.92 Negative
PC2 (Disposition): -1.00

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
weary0.254
fatigue0.249
😴0.238
sighed0.237
barely0.230
Suppressed:
C-0.268
because-0.225
B-0.219
🤟-0.217
భి-0.214