The apartment felt too quiet, the silence amplifying the thrumming in Sarah's ears. She sat on her sofa, the cushions worn and familiar but offering no comfort. The screenplay she'd been working on lay on the coffee table, its pages scattered like fallen leaves. A story that, in her opinion, was going nowhere.

Her agent's feedback had been brutal but honest. The plot was stagnant. The characters were flat. She felt like her own creative ability had evaporated. The inspiration had vanished, leaving only an empty shell.

A text message. From a contact she had made online. It simply stated: "Check out this project. You might find it interesting." She let out a humorless laugh, her fingers trembling as she picked up the phone.

Emotion: gloomy

Cluster: Sadness / Despair
PC1 (Valence): -1.68 Negative
PC2 (Disposition): -1.37

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😞0.368
无法0.319
😔0.300
いつも0.270
0.255
Suppressed:
la-0.432
l-0.419
de-0.398
B-0.294
!-0.259