Chess, Scaffolded Reasoning Laboratory
Menu

Public evidence / deterministic figure package

See the result.
Keep its boundary.

Four publication figures, exported through the same public-content boundary as the essays. Every chart carries the evidence state that determines what it can—and cannot—support.

01 / LunaTerminal frozen

Terminal frozen, exploratory, and protocol-specific.

02 / DeepSeekInterim only

Interim, unfrozen, and provisional.

03 / MechanismNot yet causal

No causal candidate-count or cross-model claim.

The visual record

Each SVG is deterministic, checksum-pinned, and provider-free. Captions and assets are copied from a fixed allowlist; operational files never enter the site.

01

Luna / terminal evidence

A completed protocol curve

Terminal · frozen
Frozen Luna joint rating estimates with profile 95 percent intervals for regular, top-10, top-5, top-3, and top-2 variants.
luna-terminal-rating-curve. Frozen Luna joint rating estimates with profile 95 percent intervals for regular, top-10, top-5, top-3, and top-2 variants. Frozen Davidson-WDL estimates. Exploratory, Black-only, adaptive-anchor protocol; regular is not k=0 on one treatment curve.
Reading boundary

These estimates describe five complete agent conditions under one Black-only adaptive ladder. Regular is a different interface condition—not k=0 on a single causal dose-response curve.

Artifact identity

Deterministic SVG · no network or provider calls · SHA-256 211baf118e2ffa15a1954b9d31d975bd976e3fe2a776be20fae9d759d1044b9e

02

DeepSeek / interim evidence

A curve still under measurement

Interim · unfrozen
Interim DeepSeek rating point estimates with working profile intervals for regular, top-10, top-5, top-3, and top-2 variants. Every value is provisional and unfrozen.
deepseek-interim-rating-curve. Interim DeepSeek rating point estimates with working profile intervals for regular, top-10, top-5, top-3, and top-2 variants. Every value is provisional and unfrozen. Working intervals; adaptive anchors and repeated looks are not corrected. No terminal DeepSeek ranking or cross-model claim.
Reading boundary

The points are a dated working synthesis. Repeated looks and adaptive anchors are not corrected, so this is neither a terminal ranking nor a cross-model comparison.

Artifact identity

Deterministic SVG · no network or provider calls · SHA-256 4d6267e4b50f38a8b86e90fe5a26f06ccff64a5d05047c6cb00c7f7bf2547b73

03

Runtime / accounting lenses

Four clocks, not four parts

Evidence boundary
Horizontal bars compare suite-worker hours, reported reply hours, active wall hours, and five-way-equivalent hours. These accounting quantities overlap and must not be added.
deepseek-runtime-accounting. Horizontal bars compare suite-worker hours, reported reply hours, active wall hours, and five-way-equivalent hours. These accounting quantities overlap and must not be added. Accounting categories are alternative lenses, not additive components; worker-hours are not elapsed human wait time.
Reading boundary

Suite-worker, reported reply, active wall, and five-way-equivalent hours overlap. Read each as an alternative denominator; adding them would count the same waiting time more than once.

Artifact identity

Deterministic SVG · no network or provider calls · SHA-256 901988f21128c2cdd56264d1f09d3e2a4b84878767e14a3bbf55a9fa67895f8c

04

Inference / gate pipeline

A claim advances only with its evidence

Evidence boundary
Three evidence lanes show that frozen Luna results support protocol-specific description, unfinished DeepSeek results require terminal freezing, and causal candidate-count claims require a paired nested-set experiment with independent evaluation.
evidence-gate-pipeline. Three evidence lanes show that frozen Luna results support protocol-specific description, unfinished DeepSeek results require terminal freezing, and causal candidate-count claims require a paired nested-set experiment with independent evaluation. A leaderboard pattern is evidence to design a causal test; it is not itself a causal result.
Reading boundary

Frozen ladder evidence supports protocol description. A candidate-count mechanism claim still requires nested candidate sets, paired positions, and an independent evaluator.

Artifact identity

Deterministic SVG · no network or provider calls · SHA-256 027abc4ecabd13c8c96bdd1115e9c6427e07460146b37172a9216bbe3d1ec171