Open method · Version 2026-08-18.1

Methodology and Source Ledger

This page makes the decision procedure inspectable. It defines what each answer means, which rule wins, why a result can be blocked, how advisories are attached, and where every material claim comes from.

Published
Evidence check
Publisher
Pharos Production
Review boundary
Independent human review not claimed

Purpose and scope

Classify the next experiment, not the permanent architecture

In scope

The method applies to language-model workloads where a team is deciding whether the current gap is external knowledge access, specialized model behavior, both, or neither. It covers retrieval data, labeled examples, evaluation, provenance, lexical and semantic retrieval, reranking, fine-tuning, PEFT/LoRA, privacy topology, runtime measurement, and operations.

Out of scope

The result does not select a vendor, model, vector store, embedding model, reranker, hosting region, security control set, legal basis, budget, or acceptance threshold. It does not certify production readiness, factual accuracy, privacy, compliance, latency, cost, or business value.

Interpretation rule: every output is a candidate for a bounded, comparable test. `BLOCKED` and `INSUFFICIENT_EVIDENCE` are valid outcomes. The engine does not convert qualitative answers into an empirical probability.

Decision procedure

Validation, precedence, then advisory tests

The pure engine reads only versioned registries and the twelve answers. It contains no hidden weights or provider defaults.

  1. Validate the record

    Each registered criterion must have one answer: Yes, No, or Unknown. Unknown IDs, missing criteria, invalid values, malformed rules, and duplicate priorities fail with stable error codes.

  2. Apply decision rules

    Decision rules are evaluated in ascending, unique priority. The first matching rule selects one outcome. Hard dependencies and missing evidence therefore run before candidate rules.

  3. Attach advisories

    Advisory rules can add lexical retrieval, citation-faithfulness, runtime, privacy, or hybrid-lifecycle tests. They never change the winning outcome.

  4. Export provenance

    The record retains answers, outcome, winning and advisory rule IDs, blockers, rejected options, unresolved inputs, next steps, source IDs, claim IDs, and registry versions.

Input dictionary

What the twelve questions mean

A Yes answer should point to an artifact or operating capability. Unknown is not a soft No; it marks evidence that can change the result.

Criteria registry 2026-08-18.1. Canonical machine-readable file: data/criteria.json.
Criterion IDEvidence questionA defensible Yes requiresWhy it changes the decision
fresh_factsMust facts change independently of a model release?Named sources, update events, freshness owner, and tolerated staleness.Changing knowledge favors an external update path rather than facts stored only in parameters.
source_citationsMust users inspect passages behind material claims?Claim-level evidence requirement, preserved source IDs, and a faithfulness review method.Retrieval can carry provenance; tuning alone does not reconstruct it.
behavior_changeIs the primary gap a stable task behavior?Observable failure class and desired behavior for format, style, classification, extraction, or instruction following.A trained adaptation needs a behavior target distinct from missing knowledge.
representative_examplesAre permitted examples available?Realistic inputs, desired outputs, failure cases, data rights, coverage, and a held-out split.Fine-tuning without representative examples is not recommended.
evaluation_setCan alternatives be compared?Versioned tasks, failure classes, baseline outputs, acceptance criteria, and review policy.No candidate can win without a shared evaluation contract.
authoritative_corpusDoes a retrieval source of truth exist?Content owners, permissions, stable IDs, metadata, updates, effective dates, and deletion semantics.RAG or hybrid is blocked when current or cited facts lack a governed corpus.
exact_term_retrievalDo exact identifiers or phrases matter?Representative exact-term queries and known semantic-retrieval misses.Adds a lexical/BM25, fusion, and reranking advisory test.
strict_runtime_budgetIs there a measured latency or request-cost ceiling?Defined traffic shape, quality target, percentile or unit, and test environment.Adds an end-to-end benchmark; it does not bias toward either label.
privacy_topologyAre data paths mapped?Storage, ingestion, retrieval, training, inference, logging, retention, access, region, and deletion paths for chosen services.Architecture labels alone do not establish privacy or retention behavior.
retrieval_operationsCan the retrieval lifecycle be operated?Owners and runbooks for ingestion, indexing, evaluation, ACL changes, stale-state detection, rollback, and deletion.A knowledge requirement is blocked when no retrieval operating path exists.
training_operationsCan the model lifecycle be operated?Owners for datasets, provider/model availability, training, evaluation, releases, regressions, rollback, and deprecation.A tuning or hybrid candidate needs an actual release capability.
stable_narrow_taskIs the trained behavior maintainable?A bounded task definition and evidence that desired behavior is not changing faster than the model release process.Broad or moving targets are better tested with prompts and workflows first.

Rule precedence

Why one result wins

Lower priority numbers run first. The order protects hard dependencies and evidence gates from being overridden by an attractive candidate pattern.

Decision rules in evaluation order. Advisory rules run after the winner and cannot change it.
PriorityRule IDWhen it matchesOutcomeDecision consequence
10RULE-BLOCKED-NO-CORPUSFresh facts or citations are required; authoritative corpus is No.BlockedReject RAG, hybrid, and tuning-as-knowledge-store. Establish the governed source first.
20RULE-INSUFFICIENT-UNKNOWN-COREAny core requirement, data asset, eval, or operating answer is Unknown.Insufficient evidenceAssign an evidence owner and resolve the unknown before choosing.
30RULE-INSUFFICIENT-NO-EVALEvaluation set is No.Insufficient evidenceCreate representative tasks and acceptance criteria.
40RULE-BASELINE-NO-EXAMPLESBehavior change is Yes; representative examples are No.Baseline firstPrompt and collect permitted success/failure examples before training.
50RULE-BLOCKED-NO-RETRIEVAL-OPSKnowledge or citations are required; corpus exists; retrieval operations are No.BlockedAssign the ingestion, evaluation, ACL, freshness, and rollback lifecycle.
60RULE-BLOCKED-NO-TRAINING-OPSBehavior and examples are Yes; training operations are No.BlockedAssign model-data, evaluation, provider, release, and rollback ownership.
70RULE-HYBRID-CANDIDATEKnowledge need and behavior need both exist; both assets and operating paths exist.Hybrid candidateTest retrieval and behavior adaptation separately before combining.
80RULE-RAG-CANDIDATEKnowledge or citations are required; corpus and retrieval operations exist.RAG candidateBuild and evaluate a retrieval baseline with provenance and stale-state tests.
90RULE-FINE-TUNING-CANDIDATEStable behavior, examples, and training operations exist; changing facts and citations are not required.Fine-tuning candidateCompare prompt-only and adapted models on task quality and regressions.
100RULE-BASELINE-NO-PRIMARY-GAPChanging facts, citations, and behavior change are all No.Baseline firstDocument a concrete failure before adding architecture.
110RULE-INSUFFICIENT-FALLBACKValidated answers do not match another candidate.Insufficient evidenceRecord the requirement or capability that prevents a match.

Advisory rules

Exact-term retrieval

ADVISORY-LEXICAL-RETRIEVAL adds a BM25 or lexical path, signal fusion, and reranking comparison when codes, names, or exact phrases matter.

Claim-level citations

ADVISORY-CITATION-FAITHFULNESS requires each material generated claim to be checked against the cited passage.

Runtime budget

ADVISORY-RUNTIME-BENCHMARK requires end-to-end latency and cost measurement for the actual provider, corpus, traffic shape, and quality target.

Mapped privacy

ADVISORY-PRIVACY-TOPOLOGY preserves the verified data-flow map and requires provider controls to be rechecked before release.

Unresolved privacy

ADVISORY-PRIVACY-UNRESOLVED adds storage, retrieval, training, logging, retention, access, and deletion mapping.

Dual lifecycle

ADVISORY-HYBRID-OPERATIONS binds corpus, index, dataset, base model, adapter or tuned model, and evaluation suite as separate release inputs.

Claim ledger

Material statements and their boundaries

`VERIFIED` means the statement follows directly from the named source within the stated scope. `BOUNDED` means the rule is a conservative synthesis whose limits are explicit.

Fifteen claims used by the decision and advisory rules. Full mappings are in data/claims.json.
Claim IDStatusBounded statementSource IDs
CLAIM-RAG-RUNTIME-RETRIEVALVerifiedRAG supplies retrieved external information at inference time rather than encoding every changing fact in model parameters.RAG paper; Microsoft
CLAIM-RAG-FRESHNESS-FITBoundedRetrieval is a suitable candidate when required knowledge changes independently and the corpus can be maintained.Microsoft; practitioner guide
CLAIM-RAG-PROVENANCEBoundedRetrieved passages can carry provenance; claim-to-passage support still needs testing.RAG paper; Anthropic
CLAIM-FINE-TUNING-BEHAVIORVerifiedFine-tuning is a path for task behavior, format, or specialization, not continuous fact synchronization.Microsoft; OpenAI
CLAIM-EXAMPLES-REQUIREDVerifiedA tuning decision needs representative examples and an evaluation set for the intended task.OpenAI; Microsoft
CLAIM-EVAL-FIRSTBoundedPrompting, retrieval, and training changes should be compared against a defined baseline.OpenAI; practitioner guide
CLAIM-LEXICAL-SEMANTICVerifiedLexical and embedding retrieval recover different signals and can be combined when both matter.Anthropic
CLAIM-RERANKINGVerifiedA reranker can reorder an initial candidate set before context reaches the model.Anthropic
CLAIM-CORPUS-QUALITYBoundedA RAG recommendation presumes an authoritative, permissioned corpus and an update path.Microsoft; practitioner guide
CLAIM-LORA-METHODVerifiedLoRA freezes pretrained weights and learns low-rank update matrices.LoRA paper
CLAIM-HYBRID-PATTERNBoundedRetrieval and adaptation can be combined when evidence and behavior are independent needs.RAFT; Microsoft
CLAIM-RAFT-SCOPEVerifiedRAFT is one retrieval-aware tuning method using relevant and distractor documents in domain-specific RAG settings.RAFT paper
CLAIM-DUAL-OPERATIONSBoundedA hybrid creates retrieval-index and model-version lifecycle work.RAFT; practitioner guide
CLAIM-PRIVACY-TOPOLOGYBoundedPrivacy and retention depend on provider, endpoint, storage, deployment, and configuration.OpenAI data controls; Microsoft
CLAIM-RUNTIME-MEASUREMENTBoundedLatency, quality, and cost tradeoffs must be measured on the target workload.Microsoft; OpenAI

Primary-source ledger

What each source is allowed to support

All eight URLs were checked on 18 August 2026. Research results and vendor statements remain scoped to their documented settings; they are not generalized into universal performance guarantees.

  1. SOURCE-PHAROS-GUIDE: practitioner framing only. Quantitative statements were not reused without primary support. The canonical URL is recorded in the source registry.
  2. SOURCE-RAG-PAPER: Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks. Used for the RAG architecture definition and provenance-oriented framing, not universal production outcomes.
  3. SOURCE-MICROSOFT-RAG-VS-FT: Augment language models with RAG or fine-tuning. Used for qualitative selection factors; examples are not treated as thresholds.
  4. SOURCE-ANTHROPIC-CONTEXTUAL-RETRIEVAL: Introducing Contextual Retrieval. Used for lexical/semantic retrieval and reranking patterns; experimental outcomes stay provider-specific.
  5. SOURCE-OPENAI-MODEL-OPTIMIZATION: Model optimization. Used for an eval-first optimization loop and representative-example requirements; product availability can change.
  6. SOURCE-LORA-PAPER: LoRA: Low-Rank Adaptation of Large Language Models. Used for the method definition; benchmarks are not generalized.
  7. SOURCE-RAFT-PAPER: RAFT: Adapting Language Model to Domain Specific RAG. Used for one retrieval-aware tuning pattern, not a claim that hybrid always wins.
  8. SOURCE-OPENAI-DATA-CONTROLS: Data controls in the OpenAI platform. Used only to show that service and endpoint configuration determine treatment; it is not a universal privacy statement.

Repository data files preserve titles, publishers, type, publication or update date when available, verification status, checked date, URL, and use boundary. The reachability checker is non-semantic: an HTTP response does not prove a claim.

Limitations

Known boundaries and change risks

No universal numeric thresholds

The method contains no generic dataset-size, document-count, accuracy, latency, token, cost, compute-saving, or hybrid-uplift threshold. Teams must set workload-specific acceptance criteria before seeing candidate results.

No automatic provenance guarantee

RAG can preserve document and chunk IDs, but citation generation and claim support are separate system behaviors. A citation-present rate is not a faithfulness measure.

No architecture-level privacy guarantee

Corpus storage, embedding services, retrieval, prompts, inference, fine-tuning uploads, logs, evaluation systems, retention, regions, access, and deletion all affect the boundary.

Provider and model drift

Fine-tuning availability, supported models, endpoint policies, pricing, context limits, retrieval services, and data controls can change. Recheck the live provider contract for every release.

Qualitative rules

Yes, No, and Unknown expose evidence gaps but compress detail. Use the exported record as an architecture-conversation input, then attach the underlying artifacts and measurements.

Review status

Research synthesis and implementation were AI-assisted and deterministically checked against the registries. Independent human technical review is not claimed. The publisher remains responsible for corrections and future versions.

Version and change policy

Update history

Public changes to method, claims, data, or interface.
DateVersionChangeEvidence action
2026-08-18.1 / app 1.0.0Initial release: criteria, ordered rules, advisories, six outcomes, examples, exports, semantic article, open data, and original diagrams.All source URLs checked; unsupported universal thresholds excluded; claim and rule references validated.

Correction request

Email info@pharosproduction.com with the page URL, claim or rule ID, a primary contrary source, and the proposed correction. A material accepted change updates the registry version, visible page, changelog, tests, and source-check date.

Reproducibility

The dependency-free source, registries, tests, and workflow are available in the public GitHub repository. Canonical data and public data copies are checked byte for byte before deployment.