tired - Story 151413 | Emotion Vectors in Gemma4-31B

The train rattled and groaned, mirroring the ache in Marcus's bones. He stared out the window at the passing landscape, each passing tree a blur. He felt as though he could sleep for a thousand years. He hadn't slept well in days. He rubbed his eyes. The constant pressure of deadlines and obligations had finally caught up to him. He was a hair's breadth from burnout. The wallet appeared on the seat beside him.

He picked it up, expecting it to be empty. But it felt heavy. Opening it, he found stacks of cash. The temptation to just keep the money and disappear was incredibly strong.

He rested his head against the cold glass of the window, trying to organize his racing thoughts. Was this a sign? A way out?

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:

`weary`	0.254
`fatigue`	0.249
`😴`	0.238
`sighed`	0.237
`barely`	0.230

Suppressed:

`C`	-0.268
`because`	-0.225
`B`	-0.219
`🤟`	-0.217
`భి`	-0.214