finding
active
finding:nla-supporting-clues-in-47-of-positive-vs-30-of-negative-claims-0-60-auroc-vs-0-77-transcript-only-adding-0-001NLA supporting clues in 47% of positive vs 30% of negative claims, 0.60 AUROC vs 0.77 transcript-only, adding +0.001
Soft clues redundant with transcript.
Source paper
extracted_from(2026) · Adam Karvonen · Euan Ong · Subhash Kantamneni · Samuel Marks
Neighborhood — ranked by edge-count
Papers (1)
paper
Claims (1)
claim
- Mechanism for null result.
Related by similarity (8)
cosine ≥ 0.65 · no typed edgeEntities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.
- Qualitative win.
- Explicit 'X causes Y' statement appears in only 0.6% (3 of 479) of positive claims' NLA outputsfinding0.797Tools rarely state causal relationships.
- Tool-induced answer flips mostly toward 'no effect' (NLA 82% positive, 84% negative), cancelling in accuracyfinding0.793Direction of tool-driven prediction changes.
- Key limitation identified: NLAs hallucinate specific details while preserving thematic accuracy; informs practical usage.
- While NLA claims can be false in specifics, they are typically thematically faithful to contextclaim0.779Key insight about confabulation patterns in NLAs enabling practical use.
- Tool-discovered hypotheses are easily guessed.
- Appendix B.
- Core limitation and usage heuristic: read NLAs for themes rather than individual factual claims; cross-check with original context.