finding
active
finding:topic-gated-poetic-shaping-gains-40-elo-on-creative-writing-bench-v3-with-no-capability-regression-0-07ppTopic-gated poetic shaping gains +40 Elo on Creative Writing Bench v3 with no capability regression (+0.07pp)
App. I.
Source paper
extracted_from(2026) · Leon Bergen · Usha Bhalla · Sidharth Baskaran · Max Loeffler +13
Neighborhood — ranked by edge-count
Papers (1)
paper
Claims (1)
claim
- Playful and poetic amplification support.
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.
- Case study demonstrating mechanism behind flat harness-updating: smaller models reach same procedural content
- Full evolver-side SWE results showing comparable performance across Claude family tiers
- Case demonstrating that model scale does not predict harness-updating quality
- Battery does not detect beautiful writing; poetic prompt boosts aesthetics while suppressing self-observation
- Robustness result showing introspection stage's contribution for Llama model
- VS increases diversity by 1.6-2.1x over direct prompting on creative writing tasks (poem, story, joke)finding0.722Core empirical result demonstrating VS's effectiveness on creative writing diversity
- Quantifies harness activation failure for weak-tier models vs. strong-tier models
- Complementary temporal activation pattern suggesting distinct roles for OTD and backtracking latent classes