"Well," I said aloud, the email still open on my laptop screen. The midday sun streamed through the window, illuminating the dust motes dancing in the air. "That's that, then." I closed the email and took a long, exaggerated stretch, my back cracking pleasurably. The rejection? Almost expected, really.

I walked over to the bookshelf, my gaze lingering on a particularly handsome first edition, its gold lettering glinting in the light. A true masterpiece, just like my manuscript. But it would be wasted on those philistines. Clearly, they couldn’t recognize genius when it slapped them in the face. I poured myself a generous glass of vintage champagne, the bubbles fizzing deliciously.

The telephone rang. It was my agent, her voice laced with a forced cheerfulness. I cut her off. "Don't worry," I said, my tone smooth and unwavering. "We'll find a publisher who understands." I smiled, a small, knowing curve of my lips. It wouldn’t take long. They’d be begging to publish me soon enough.

Emotion: smug

Cluster: Positive / Joy
PC1 (Valence): 3.43 Positive
PC2 (Disposition): 1.71

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
de1.109
l0.615
la0.585
😎0.570
😉0.564
Suppressed:
S-0.579
own-0.486
💔-0.483
😞-0.464
😣-0.455