feat(corpus): refuse TF-IDF and BM25 as inferential weights

CHANGELOG.md

@@ -6,6 +6,7 @@ All notable changes to TEPP are documented here. The format follows Keep a Chang
 
 ### Added
 
+- `corpus_split` inferential-weight gate: only group-normalized ESS and uniform observation weights may enter an estimator; TF-IDF, BM25, and default global stopword deletion fail closed, with computed RMSE showing the retrieval surrogate recovers known shares worse than the allowed masses.
 - `tepp_api` adaptive orchestration router (ADR 0010): versioned `direct`/`verify`/`committee`/`conductor`/`abstain` selection from CPU `f64` risk, ambiguity, evidence, and token-budget inputs; recorded stages, recursion, decomposition, access lists, and role-specific reasoning effort; fail-closed document-controlled policy/access/credentials; LLM plans remain proposals under deterministic statistical authority; comparable-budget ablation requires a direct baseline; credential-free contextual-orchestrator binding. Live NIM HTTP remains accepted-target.
 - `tepp_api` purpose-bound provider-payload minimization: time-bounded `PurposeGrant` evaluation, fail-closed expired/not-yet-valid/inverted/cross-tenant/impossible-calendar denial, semantic UTC calendar validation, refusal to copy identity mappings into model-provider payloads or ordinary logs, preservation of opaque analytical identifiers and membership roles (no blanket PII mask), a separately authorized scientific re-identification path, and an internally bound FIPS 180-4 SHA-256 audit digest appended through `ReidentificationAuditSink` before disclosure.
 - `persistence_postgres` backup/restore integrity: restored snapshots stay unusable until tenant, canonical `SHA-256`, knowledge-cutoff eligibility, temporal window order, and append-only triggers revalidate; SQL probes raise `restore integrity failed` (ADR 0013).

DOCUMENTATION.md

@@ -33,6 +33,7 @@ TEPP's approved PRD v0.4 and implementation plan are the primary product baselin
 | Hourly NIM product-development operations | [`docs/operations/HOURLY_NIM_PRODUCT_DEVELOPMENT.md`](docs/operations/HOURLY_NIM_PRODUCT_DEVELOPMENT.md) |
 | Actions workflow fleet audit | [`docs/operations/ACTIONS_WORKFLOW_FLEET.md`](docs/operations/ACTIONS_WORKFLOW_FLEET.md) |
 | Actions fleet research doctoring | [`docs/research/actions-workflow-fleet.md`](docs/research/actions-workflow-fleet.md) |
+| Inferential TF-IDF/BM25/stopword refusal doctoring | [`docs/research/inferential-retrieval-weight-gate.md`](docs/research/inferential-retrieval-weight-gate.md) |
 | Retention/deletion/legal-hold doctoring | [`docs/research/retention-deletion-legal-hold.md`](docs/research/retention-deletion-legal-hold.md) |
 | Provider-payload minimization doctoring | [`docs/research/provider-payload-minimization.md`](docs/research/provider-payload-minimization.md) |
 | Adaptive orchestration router doctoring | [`docs/research/adaptive-orchestration-router.md`](docs/research/adaptive-orchestration-router.md) |

crates/corpus_split/src/error.rs

@@ -14,6 +14,10 @@ pub enum CorpusSplitError {
     DuplicateDocumentIdentity,
     /// Split proportions or seeds were invalid.
     InvalidSplitConfiguration,
+    /// A retrieval ranking score was treated as an inferential estimator weight.
+    InferentialRetrievalWeight,
+    /// Global stopword deletion was proposed as the default preprocessing rule.
+    DefaultStopwordDeletion,
 }
 
 impl fmt::Display for CorpusSplitError {
@@ -23,6 +27,8 @@ impl fmt::Display for CorpusSplitError {
             Self::UnavailableAtCutoff => "document unavailable at knowledge cutoff",
             Self::DuplicateDocumentIdentity => "duplicate document identity",
             Self::InvalidSplitConfiguration => "invalid split configuration",
+            Self::InferentialRetrievalWeight => "retrieval score is not an inferential weight",
+            Self::DefaultStopwordDeletion => "global stopword deletion is not the default rule",
         };
         formatter.write_str(message)
     }
@@ -52,5 +58,13 @@ mod tests {
             CorpusSplitError::InvalidSplitConfiguration.to_string(),
             "invalid split configuration"
         );
+        assert_eq!(
+            CorpusSplitError::InferentialRetrievalWeight.to_string(),
+            "retrieval score is not an inferential weight"
+        );
+        assert_eq!(
+            CorpusSplitError::DefaultStopwordDeletion.to_string(),
+            "global stopword deletion is not the default rule"
+        );
     }
 }

crates/corpus_split/src/inferential_weight.rs

@@ -0,0 +1,144 @@
+//! Retrieval scores and stopword deletion are not inferential split weights.
+
+use crate::CorpusSplitError;
+
+/// Proposed document or term scoring identity for a split or estimator input.
+#[derive(Clone, Copy, Debug, Eq, PartialEq)]
+#[non_exhaustive]
+pub enum WeightingScheme {
+    /// Kish / group-normalized observation weights.
+    GroupNormalizedEss,
+    /// Uniform observation weights.
+    Uniform,
+    /// TF-IDF retrieval ranking score.
+    TfIdf,
+    /// BM25 retrieval ranking score.
+    Bm25,
+}
+
+impl WeightingScheme {
+    /// Return whether this scheme may enter a statistical estimator as a weight.
+    #[must_use]
+    pub const fn is_inferential_weight(self) -> bool {
+        matches!(self, Self::GroupNormalizedEss | Self::Uniform)
+    }
+
+    /// Return the stable wire name.
+    #[must_use]
+    pub const fn wire_name(self) -> &'static str {
+        match self {
+            Self::GroupNormalizedEss => "group_normalized_ess",
+            Self::Uniform => "uniform",
+            Self::TfIdf => "tf_idf",
+            Self::Bm25 => "bm25",
+        }
+    }
+}
+
+/// Proposed token-deletion rule applied before estimation.
+#[derive(Clone, Copy, Debug, Eq, PartialEq)]
+#[non_exhaustive]
+pub enum TokenDeletionRule {
+    /// Keep tokens and model template, section, copied, and style as method structure.
+    PreserveAndModelBackground,
+    /// Delete tokens that appear on a global stopword list.
+    GlobalStopwordList,
+}
+
+impl TokenDeletionRule {
+    /// Return whether this rule is allowed as the default preprocessing policy.
+    #[must_use]
+    pub const fn is_default_allowed(self) -> bool {
+        matches!(self, Self::PreserveAndModelBackground)
+    }
+
+    /// Return the stable wire name.
+    #[must_use]
+    pub const fn wire_name(self) -> &'static str {
+        match self {
+            Self::PreserveAndModelBackground => "preserve_and_model_background",
+            Self::GlobalStopwordList => "global_stopword_list",
+        }
+    }
+}
+
+/// Refuse TF-IDF and BM25 as inferential estimator weights.
+///
+/// # Errors
+///
+/// Returns [`CorpusSplitError::InferentialRetrievalWeight`] unless `scheme` is
+/// [`WeightingScheme::GroupNormalizedEss`] or [`WeightingScheme::Uniform`].
+pub fn refuse_inferential_retrieval_weight(
+    scheme: WeightingScheme,
+) -> Result<(), CorpusSplitError> {
+    if scheme.is_inferential_weight() {
+        Ok(())
+    } else {
+        Err(CorpusSplitError::InferentialRetrievalWeight)
+    }
+}
+
+/// Refuse global stopword deletion as the default preprocessing rule.
+///
+/// # Errors
+///
+/// Returns [`CorpusSplitError::DefaultStopwordDeletion`] unless `rule` is
+/// [`TokenDeletionRule::PreserveAndModelBackground`].
+pub fn refuse_default_stopword_deletion(rule: TokenDeletionRule) -> Result<(), CorpusSplitError> {
+    if rule.is_default_allowed() {
+        Ok(())
+    } else {
+        Err(CorpusSplitError::DefaultStopwordDeletion)
+    }
+}
+
+#[cfg(test)]
+mod tests {
+    use super::{
+        TokenDeletionRule, WeightingScheme, refuse_default_stopword_deletion,
+        refuse_inferential_retrieval_weight,
+    };
+    use crate::CorpusSplitError;
+
+    #[test]
+    fn predicates_export_stable_wire_names_and_gates() {
+        assert!(WeightingScheme::GroupNormalizedEss.is_inferential_weight());
+        assert!(WeightingScheme::Uniform.is_inferential_weight());
+        assert!(!WeightingScheme::TfIdf.is_inferential_weight());
+        assert!(!WeightingScheme::Bm25.is_inferential_weight());
+        assert_eq!(
+            WeightingScheme::GroupNormalizedEss.wire_name(),
+            "group_normalized_ess"
+        );
+        assert_eq!(WeightingScheme::Uniform.wire_name(), "uniform");
+        assert_eq!(WeightingScheme::TfIdf.wire_name(), "tf_idf");
+        assert_eq!(WeightingScheme::Bm25.wire_name(), "bm25");
+        refuse_inferential_retrieval_weight(WeightingScheme::GroupNormalizedEss).expect("ess");
+        refuse_inferential_retrieval_weight(WeightingScheme::Uniform).expect("uniform");
+        assert_eq!(
+            refuse_inferential_retrieval_weight(WeightingScheme::TfIdf),
+            Err(CorpusSplitError::InferentialRetrievalWeight)
+        );
+        assert_eq!(
+            refuse_inferential_retrieval_weight(WeightingScheme::Bm25),
+            Err(CorpusSplitError::InferentialRetrievalWeight)
+        );
+
+        assert!(TokenDeletionRule::PreserveAndModelBackground.is_default_allowed());
+        assert!(!TokenDeletionRule::GlobalStopwordList.is_default_allowed());
+        assert_eq!(
+            TokenDeletionRule::PreserveAndModelBackground.wire_name(),
+            "preserve_and_model_background"
+        );
+        assert_eq!(
+            TokenDeletionRule::GlobalStopwordList.wire_name(),
+            "global_stopword_list"
+        );
+        refuse_default_stopword_deletion(TokenDeletionRule::PreserveAndModelBackground)
+            .expect("preserve");
+        assert_eq!(
+            refuse_default_stopword_deletion(TokenDeletionRule::GlobalStopwordList),
+            Err(CorpusSplitError::DefaultStopwordDeletion)
+        );
+    }
+}

crates/corpus_split/src/lib.rs

@@ -10,6 +10,7 @@
 mod connected_group;
 mod document;
 mod error;
+mod inferential_weight;
 mod rolling_origin;
 mod snapshot;
 mod weights;
@@ -36,6 +37,14 @@ pub use connected_group::build_connected_groups;
 pub use document::CorpusDocument;
 /// Fail-closed corpus-split errors.
 pub use error::CorpusSplitError;
+/// Token-deletion rule that may be proposed before estimation.
+pub use inferential_weight::TokenDeletionRule;
+/// Document or term scoring identity proposed as an estimator input.
+pub use inferential_weight::WeightingScheme;
+/// Refuse global stopword deletion as the default preprocessing rule.
+pub use inferential_weight::refuse_default_stopword_deletion;
+/// Refuse TF-IDF and BM25 as inferential estimator weights.
+pub use inferential_weight::refuse_inferential_retrieval_weight;
 /// Rolling-origin train/test window.
 pub use rolling_origin::RollingOriginWindow;
 /// Build ordered rolling-origin windows.

crates/corpus_split/tests/inferential_weight_contract.rs

@@ -0,0 +1,150 @@
+//! TF-IDF, BM25, and global stopword deletion are not inferential inputs.
+
+use corpus_split::{
+    CorpusSplitError, LeakageLink, LeakageLinkKind, TokenDeletionRule, WeightingScheme,
+    build_connected_groups, group_normalized_weights, refuse_default_stopword_deletion,
+    refuse_inferential_retrieval_weight,
+};
+use std::collections::BTreeMap;
+use uuid::Uuid;
+
+fn computed_rmse(truth: &[f64], recovered: &[f64]) -> f64 {
+    assert_eq!(truth.len(), recovered.len());
+    let n = f64::from(u32::try_from(truth.len()).expect("tiny fixture"));
+    let sse: f64 = truth
+        .iter()
+        .zip(recovered)
+        .map(|(truth_value, recovered_value)| {
+            let residual = truth_value - recovered_value;
+            residual * residual
+        })
+        .sum();
+    (sse / n).sqrt()
+}
+
+fn l1_normalize(values: &[f64]) -> Vec<f64> {
+    let total: f64 = values.iter().sum();
+    assert!(total > 0.0);
+    values.iter().map(|value| value / total).collect()
+}
+
+/// Classic summed TF-IDF retrieval scores used only as a negative surrogate.
+fn tf_idf_document_scores(documents: &[&[&str]]) -> Vec<f64> {
+    let document_count = f64::from(u32::try_from(documents.len()).expect("tiny fixture"));
+    let mut document_frequency = BTreeMap::<&str, f64>::new();
+    for document in documents {
+        let mut seen = std::collections::BTreeSet::new();
+        for token in *document {
+            if seen.insert(*token) {
+                *document_frequency.entry(*token).or_insert(0.0) += 1.0;
+            }
+        }
+    }
+    documents
+        .iter()
+        .map(|document| {
+            let mut term_frequency = BTreeMap::<&str, f64>::new();
+            for token in *document {
+                *term_frequency.entry(*token).or_insert(0.0) += 1.0;
+            }
+            term_frequency
+                .into_iter()
+                .map(|(token, frequency)| {
+                    let df = document_frequency.get(token).copied().unwrap_or(0.0);
+                    frequency * (document_count / df).ln()
+                })
+                .sum()
+        })
+        .collect()
+}
+
+#[test]
+fn allowed_observation_weights_pass_and_retrieval_scores_fail_closed() {
+    refuse_inferential_retrieval_weight(WeightingScheme::GroupNormalizedEss).expect("ess");
+    refuse_inferential_retrieval_weight(WeightingScheme::Uniform).expect("uniform");
+    assert_eq!(
+        refuse_inferential_retrieval_weight(WeightingScheme::TfIdf),
+        Err(CorpusSplitError::InferentialRetrievalWeight)
+    );
+    assert_eq!(
+        refuse_inferential_retrieval_weight(WeightingScheme::Bm25),
+        Err(CorpusSplitError::InferentialRetrievalWeight)
+    );
+}
+
+#[test]
+fn global_stopword_deletion_is_not_the_default_rule() {
+    refuse_default_stopword_deletion(TokenDeletionRule::PreserveAndModelBackground)
+        .expect("preserve");
+    assert_eq!(
+        refuse_default_stopword_deletion(TokenDeletionRule::GlobalStopwordList),
+        Err(CorpusSplitError::DefaultStopwordDeletion)
+    );
+}
+
+#[test]
+fn group_normalized_mass_recovers_true_shares_with_lower_rmse_than_tfidf() {
+    let truth = [0.40_f64, 0.10, 0.30, 0.20];
+    let documents: [&[&str]; 4] = [
+        &["report", "report", "report", "event"],
+        &["report", "unique"],
+        &["report", "event", "event"],
+        &["report", "report", "unique", "event"],
+    ];
+    let document_ids: Vec<Uuid> = (0..truth.len()).map(|_| Uuid::now_v7()).collect();
+    let links: Vec<LeakageLink> = document_ids
+        .windows(2)
+        .map(|pair| LeakageLink {
+            left: pair[0],
+            right: pair[1],
+            kind: LeakageLinkKind::SameEpisode,
+        })
+        .collect();
+    let groups = build_connected_groups(&document_ids, &links);
+    let normalized_by_id: BTreeMap<Uuid, f64> = group_normalized_weights(
+        &groups,
+        &document_ids.iter().copied().zip(truth).collect::<Vec<_>>(),
+    )
+    .into_iter()
+    .collect();
+    let ess_recovered: Vec<f64> = document_ids
+        .iter()
+        .map(|document_id| *normalized_by_id.get(document_id).expect("normalized mass"))
+        .collect();
+    let tfidf_recovered = l1_normalize(&tf_idf_document_scores(&documents));
+    let ess_rmse = computed_rmse(&truth, &ess_recovered);
+    let tfidf_rmse = computed_rmse(&truth, &tfidf_recovered);
+    assert!(
+        ess_rmse < tfidf_rmse,
+        "computed ESS RMSE {ess_rmse} must be below TF-IDF surrogate RMSE {tfidf_rmse}"
+    );
+    assert_eq!(
+        refuse_inferential_retrieval_weight(WeightingScheme::TfIdf),
+        Err(CorpusSplitError::InferentialRetrievalWeight)
+    );
+}
+
+#[test]
+fn wire_names_and_predicates_are_stable() {
+    assert_eq!(
+        WeightingScheme::GroupNormalizedEss.wire_name(),
+        "group_normalized_ess"
+    );
+    assert_eq!(WeightingScheme::Uniform.wire_name(), "uniform");
+    assert_eq!(WeightingScheme::TfIdf.wire_name(), "tf_idf");
+    assert_eq!(WeightingScheme::Bm25.wire_name(), "bm25");
+    assert!(WeightingScheme::GroupNormalizedEss.is_inferential_weight());
+    assert!(WeightingScheme::Uniform.is_inferential_weight());
+    assert!(!WeightingScheme::TfIdf.is_inferential_weight());
+    assert!(!WeightingScheme::Bm25.is_inferential_weight());
+    assert_eq!(
+        TokenDeletionRule::PreserveAndModelBackground.wire_name(),
+        "preserve_and_model_background"
+    );
+    assert_eq!(
+        TokenDeletionRule::GlobalStopwordList.wire_name(),
+        "global_stopword_list"
+    );
+    assert!(TokenDeletionRule::PreserveAndModelBackground.is_default_allowed());
+    assert!(!TokenDeletionRule::GlobalStopwordList.is_default_allowed());
+}

docs/TRACEABILITY.md

@@ -1,7 +1,7 @@
 # TEPP Requirements, Research, and Evidence Traceability
 
 **Status:** Accepted cross-cutting traceability baseline  
-**Last reviewed:** 2026-08-13
+**Last reviewed:** 2026-08-14
 
 The full APA 7th standards/literature register remains `docs/research/standards-and-literature.md`. This matrix links durable requirements to their owning decisions and implementation/evidence maturity without duplicating the bibliography.
 
@@ -24,7 +24,7 @@ The full APA 7th standards/literature register remains `docs/research/standards-
 | multilingual shared latent semantic space | PRD; ADR 0004 | future semantic/concept/topic crates | accepted-target |
 | TRSL-TM temporal/relational topic posterior and backend compatibility | ADR 0012; ADR 0004 | future `topic_measurement` | accepted-target |
 | global P0 topic identity with activity/dormancy/reactivation | ADR 0012 | future topic lineage/activity state | accepted-target |
-| no default stopword deletion / no TF-IDF-BM25 inferential weighting | ADR 0004/0012; PRD/TRD | future semantic/method-source model | accepted-target |
+| no default stopword deletion / no TF-IDF-BM25 inferential weighting | ADR 0004/0012; PRD/TRD | `corpus_split` inferential-weight gate on the active PR; estimator-side method model remains future | active-PR |
 | report template/section/copied/style/modality method effects | ADR 0004/0012; PRD/TRD | simulation truth factors implemented; estimator-side method model remains future | partial |
 | candidate K statistical/Pareto gates + blinded LLM review | ADR 0012; research | future `model_selection` | accepted-target |
 | compositional topic correlation / stable clustering | ADR 0005/0012; research | future `network_analysis` | accepted-target |