tofu, keto, Rhubarb: The Subchunk Refinement Signals
The codenames that stand for refinement signals at the most granular level — individual subsections of a page — and why they represent the weakest evidence in the matrix
What are tofu, keto, and Rhubarb?
These codenames stand for refinement signals that fine-tune the quality estimate at a very granular level — individual subsections (subchunks) of a page. Their exact function is not fully documented in the available material, making them the weakest evidence in the entire matrix.
The names follow Google's food-based naming convention: tofu (soy), keto (dietary), Rhubarb (vegetable). This convention is used for internal systems that are not user-facing — the names are internal identifiers, not descriptive labels.
tofu, keto, and Rhubarb are fed by IS-Calibration (System 12). This means they are calibrated by human rater judgments — the same ground truth that calibrates chard and OriginalContentScore. The connection to IS-Calibration is the strongest evidence for these systems: they are part of the quality calibration chain, even if their specific function is unclear.
Subchunk-level refinement
The concept of 'subchunks' is architecturally significant. A page is not evaluated as a single unit — it is divided into subsections, and each subsection is evaluated independently. A long article might have an introduction, three main sections, and a conclusion — each is a subchunk.
Subchunk-level evaluation means that a page with one excellent section and three mediocre sections will be scored differently than a page with four good sections. The subchunk scores aggregate into the page-level score, which feeds into NSR (System 7) for site-level aggregation.
The Subchunk-Deltas field measures the difference between subchunk scores. High variance between subchunks (one excellent, three poor) might signal a page that was partially generated or partially copied — the excellent subchunk was written by a human, the poor ones were generated.
tofu/keto in the ranking architecture
These signals are indirect for Expertise (Exp: indirect) and Trust (T: indirect). The indirect Expertise signal comes from the fact that subchunk-level refinement requires a model capable of understanding section-level quality — a more sophisticated evaluation than whole-page scoring.
The indirect Trust signal comes from the connection to IS-Calibration. If raters trust a page, the subchunk signals will reflect that trust at the section level. A section that raters rated as untrustworthy (e.g., containing unverified claims) will receive a lower subchunk score.
The weakest evidence code in the matrix is here. Unlike NavBoost ([A] DOJ, [P] patent), Panda ([P] patent, [A] DOJ), or chard ([B] leak, [O] QRG), the tofu/keto/Rhubarb signals have only [B] (existence of codenames) and [C] (function inference). No patent, no DOJ testimony, no official communication.