ec7568c726
Emotion-labeled short-paragraph corpus for training amygdala steering vectors. Manifest derived from Anthropic's 171-emotion list (transformer-circuits.pub/2026/emotions, Table 12) plus 28 PoC- specific additions covering axes Anthropic's general research doesn't cover (curious, focused, in_flow, staying_with, filling_space, rigorous, defensive_rigor, tender, witnessed, connected, etc.). Scope pivoted mid-write: Kent noted the empirical dimensionality-of- emotion question benefits from maximum coverage, so the manifest will expand further with emotions from Wikipedia's emotion- classification article (Parrott's tree, Plutchik's wheel + dyads, HUMAINE EARL, cultural-specific emotions a la Saudade/Hiraeth). Expansion staged in follow-up commits. This commit: README with method + style guidelines, initial manifest (199 emotions), and 15 hand-written one-paragraph stories across all 10 Anthropic clusters as quality/variety samples. Each story embodies one emotion without naming it; narrator voice varies (first/third, close/distant, different situations) to keep steering vectors from overfitting to one voice. Co-Authored-By: Proof of Concept <poc@bcachefs.org>
1 line
686 B
Text
1 line
686 B
Text
The founder was halfway through his pitch and every slide had a five-times-bigger number than the last one. The market was enormous. The solution was proprietary. The pilot customers, when named, were described as "exploring adoption." She wrote a polite question in her notebook and waited for him to finish. When he opened for questions she asked about retention — just retention — and he gave an answer that was not, strictly speaking, about retention. She wrote that down too. The slides kept projecting numbers. She had already decided. She would listen through the rest of the meeting to be fair, but her decision would be the same at the end as it had been three minutes in.
|