The coffee tasted like ash in Michael’s mouth. Across the fence, Mrs. Gable was pointing at the sprawling maple. Her smile, tight and brittle, didn’t reach her eyes. "It's casting such a shadow, dear. And the leaves... they're a nightmare." Michael felt a clench in his stomach. He didn't want to engage in conversation, in a discussion about anything.

He took another sip and found he didn't even want to move. The sun beat down on the parched lawn, and he could feel the moisture draining from him. It was a suffocating pressure, the weight of the day, of the past few weeks, all bearing down. Her suggestion, however innocent, felt like a deliberate jab, twisting the knife. He nodded numbly, unable to form a coherent reply, and the knot in his chest tightened.

Emotion: troubled

Cluster: Fear / Anxiety
PC1 (Valence): -2.20 Negative
PC2 (Disposition): -0.81

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😞0.343
S0.317
ness0.302
😔0.294
😢0.272
Suppressed:
de-0.553
la-0.511
l-0.307
/-0.260
(!)-0.236