Liam was perched on his bed, the afternoon sun slanting through his window, illuminating dust motes dancing in the air. He was lost in the echo of a forgotten song when he saw it: Mrs. Davis, his English teacher, had started following him on Twitter. He'd never expected it. He closed his eyes, remembering the way she always encouraged his writing, the way her eyes crinkled when she laughed.

His throat felt tight. He suddenly remembered a story he’d written about his grandfather, a man he'd dearly loved. He wanted to share it with her now, to somehow express a feeling that swelled within him, a tender longing for the past.

He grabbed his laptop, hesitating before he opened the document, the cursor blinking on a blank page. The simple act of her following him unlocked something deep inside, a yearning for connection he hadn't known was there. He wrote a note to himself, “Tell her about Grandpa.”

Emotion: sentimental

Cluster: Nostalgia / Reflection
PC1 (Valence): 0.89 Positive
PC2 (Disposition): -1.58

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
L0.677
fondly0.567
memories0.503
0.488
reminis0.466
Suppressed:
either-0.515
or-0.488
[]-0.473
--0.408
не-0.396