# DEEPREAD: DEEPREAD

> Living preprint · as of 2026-10-25 · map node X-040 (experiment) · interim record, not a final result.

<!-- lint:exempt:start -->
- [X-040] DEEPREAD · DEEPREAD · question: Does the deep model (Qwen3.6-35B-A3B UD-Q2_K_XL) convert gold retrieved passages (R450 oracle contexts) better than the 4B (Qwen3-4B-Instruct-2507) reading the identical contexts, on v0.1 EN 57? · result: +0.029 (−0.004…+0.063) · set v0.1 EN 57, judge Claude Opus, n 57, seeds 1 (seed 42), CI −0.004…+0.063 · verdict: not-detected (INCONCLUSIVE (frozen band 0.606 ≤ x < 0.660; deep35 R450 0.6345); CLOSED R15 records it as 'not detected at MDE 0.055')

<!-- lint:exempt:end -->
## Outgoing

- ruled out: [G-cut] cut (frozen build) (more 4B builds did not move held-out quality beyond noise) → https://r4topunk.github.io/boar-research/map/g-cut.md

Read chapter 6, The deep tier: https://r4topunk.github.io/boar-research/research/deep-tier.md

## Facts cited {#facts}

Resolve every [F-…] marker in https://r4topunk.github.io/boar-research/data/facts.json. Each line gives the claim, the number with its metric family, set and status, judge, n, seeds, CI, date and caveat.

<!-- lint:exempt:start -->
- [F-522] R15 Deep reader advantage: gold passages. +0.029 (−0.004…+0.063) · set UNKNOWN (status UNKNOWN) · judge UNKNOWN · n UNKNOWN · seeds UNKNOWN · CI not computed · caveat: why closed: not detected at MDE 0.055

<!-- lint:exempt:end -->
---

Cite as: r4to, independent researcher. RatLab — A Boar Research (2026), research record 1.0, as of 2026-10-25. Text and data CC BY 4.0. Fact ids or a link must travel with a quote. How to cite: https://r4topunk.github.io/boar-research/cite/ · https://r4topunk.github.io/boar-research/CITATION.cff.

AI-reviewed, not clinically reviewed. Nothing here is medical advice.