Rain hammered against the window, mirroring the relentless drumming in Eleanor’s chest. Another rejection email. Another day closer to eviction. She slumped further onto the threadbare couch, the cushions offering no comfort. Her phone pinged. A notification. Hope flickered, only to be extinguished when she saw the sender: David. Not the David she was expecting. This was David, her former colleague, from years ago. The text read: “Just saw your LinkedIn. Glad to see you’re still… around.” Eleanor felt a hot flush crawl up her neck.

She stared at the message, re-reading it, dissecting each carefully chosen word. Around. As if she was a piece of furniture, occupying space but serving no purpose. Her fingers, trembling, typed a reply, then deleted it. What was the point? They weren't friends. They hadn't spoken in ages. The phone slipped from her grasp and landed with a soft thud on the floor.

She closed her eyes, imagining all the things she *could* have been. A successful businesswoman, a loving partner, a homeowner. Now, just *around*. She hated the thought of her life being a disappointment to others, but she felt like the biggest disappointment to herself.

Emotion: worthless

Cluster: Sadness / Despair
PC1 (Valence): -2.10 Negative
PC2 (Disposition): -0.15

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
L0.427
S0.405
😞0.386
ness0.339
缺乏0.326
Suppressed:
de-0.320
(!)-0.286
随后-0.268
😍-0.262
tc-0.255