#U.LanguageStateRepresentationFactorBundle

Pattern C.2.7 · Draft · Definitional (D) · Normative unless marked informative Part C - Kernel Extension Specifications

Type: Definitional (D) Status: Draft Normativity: Normative unless marked informative

Plain-name. Language-state representation-factor bundle.

Published position claims in the declared language-state chart over U.CharacteristicSpace must distinguish representation factors such as locality, sparsity, and symbolicity without pretending they form one master axis.

#Keywords

• representation factors
• locality
• sparsity
• symbolicity
• factor bundle
• representation organization.

#Relations

#Content

#Problem frame

Published position claims in the declared language-state chart over U.CharacteristicSpace must distinguish representation factors such as locality, sparsity, and symbolicity without pretending they form one master axis.

#Problem

Terms such as EncodingBasis collapse several independent choices. That makes comparison brittle and encourages one-dimensional stories such as distributed = informal or local = precise.

#Forces

#Solution

U.LanguageStateRepresentationFactorBundle is a factor bundle, not one scalar characteristic. The minimal core starter set is:

• U.LocalityDistribution
• U.Sparsity
• U.Symbolicity

A Context may publish a local alias such as EncodingBasis, but it shall dock back to the underlying factor bundle instead of replacing it.

#Minimal factor readings

#Non-collapse rules

LanguageStateRepresentationFactorBundle is not:

• LanguageStateAnchoringMode;
• Formality;
• ArticulationExplicitness;
• LanguageStateClosureDegree.

A representation may be distributed yet strongly trace-anchored; symbolic yet weakly articulated; sparse yet low-closure. Those combinations shall remain visible.

#Extension rule

Contexts may add extra representation factors only if the extension is published as a factor addition rather than as a new master axis that erases the core factor bundle.

#Archetypal Grounding

Tell. A model-state cue can be highly distributed but still strongly trace-anchored; a symbolic note can be low articulation if its semantics are still vague.

Show (System). An operator decision aid may mix sparse alert codes and symbolic procedure text.

Show (Episteme). A research probe can move from distributed activation patterns to sparse symbolic hypotheses without any one-step formality story.

#Bias-Annotation

The pattern resists folk theories that try to line up one representation axis with one stage or progression story.

#Conformance Checklist

• CC-C.2.7-1 LanguageStateRepresentationFactorBundle SHALL be published as a factor bundle, not as a hidden scalar.
• CC-C.2.7-2 Local aliases such as EncodingBasis MAY exist only with an explicit docking to the owned factors.
• CC-C.2.7-3 Representation factors MUST NOT silently replace LanguageStateAnchoringMode or LanguageStateClosureDegree.
• CC-C.2.7-4 New local factors SHALL preserve the factor-bundle discipline.

#Common Anti-Patterns and How to Avoid Them

• One-axis myth. Treating distributed/local or symbolic/subsymbolic as the whole story.
• Progression collapse. Equating representation shifts with formalization or closure.
• Alias capture. Letting EncodingBasis or a similar local alias erase the factor bundle.

#Consequences

The benefit is cleaner comparison across schools, substrates, and publication forms. The trade-off is that representation talk becomes more explicit and less slogan-friendly.

#Rationale

The factor-bundle design keeps the representation basis-slot family in the declared language-state chart over U.CharacteristicSpace orthogonal to articulation, closure, and anchoring.

#SoTA-Echoing

This factorization fits current work on sparse distributed representations, symbolic/neuro-symbolic stacks, and interpretability practice.

#Relations

• Builds on: A.18, C.2.2a, C.2.LS.
• Coordinates with: C.2.6, A.16.0, A.16, A.16.1, B.4.1, B.5.2.0, F.9.1.
• Constrains: language-state position publication and bridge loss notes around representation shifts.

#Worked Examples and Factor Interaction Notes

#Distributed but explicit

A model-side summary may be representation-wise distributed and still highly explicit once published into a stable symbolic wrapper. This case matters because it blocks the folk myth that distributed implies vague.

#Symbolic but still weakly articulated

A glossary-like note may be fully symbolic while still low in AE because the semantic anchors are not yet stabilized. This blocks the opposite myth: symbolic therefore explicit.

#Mixed-stack publication

An operator-facing publication face may combine sparse alert codes, symbolic procedure text, and distributed back-end model summaries. The representation-factor bundle should make that mixture visible instead of compressing it into one label.

#Authoring and Review Guidance

#Author prompt

To publish a representation-factor bundle, ask separately:

• how local or distributed is the representation?
• how sparse or dense is it?
• how symbolic or subsymbolic is it?
• which additional factor, if any, genuinely matters enough to publish?

#Review prompt

A reviewer should reject any attempt to use one factor as if it summarized the rest. The factor bundle exists precisely to block that reduction.

#Cross-facet reminder

Reviewers should also watch for silent replacement of LanguageStateAnchoringMode, AE, or CD by representation talk.

#Extension and Migration Notes

#Local extension rule

Contexts may add extra factors, but each added factor should answer a distinct question rather than duplicating locality, sparsity, or symbolicity under another label.

#Migration from alias-heavy prose

Aliases such as EncodingBasis or similar should be unfolded into explicit factor dockings before they are relied upon for routing, comparison, or bridge claims.

#Boundary reminder

U.LanguageStateRepresentationFactorBundle describes representational organization only. It does not determine route authority, closure, or anchoring by itself.

#Factor-Bundle Publication Discipline

#Minimal representation package

A publishable U.LanguageStateRepresentationFactorBundle should normally show the current factor settings for locality/distribution, sparsity/density, and symbolicity/subsymbolicity, together with any declared extra factor. If a factor is intentionally omitted, say so rather than hiding the omission under a compact alias.

#No hidden scalar rule

Compact overlays such as "sparse-symbolic" are lawful only when they dock to the underlying factor bundle. No compact label may behave as a hidden master score for routing, bridge comparison, or stage/progression talk.

#Alias docking rule

Local aliases such as EncodingBasis are lawful only when their docking to the owned factors is explicit and stable. If an alias compresses several factors, the compression should remain visible.

#Factor Interaction and Cross-Facet Reading Law

#Interaction law

Representation factors may correlate, but they do not determine one another. Highly distributed cues can still be sparse; symbolic publications can still be locally dense; mixed symbolicity can coexist with either strong or weak articulation. Publish the actual factor bundle rather than narrating one factor as if it predicted the rest.

#Cross-facet non-substitution

Representation talk must not silently replace AE, CD, or LanguageStateAnchoringMode. A shift from distributed to symbolic publication may change readability while leaving articulation low, closure open, or anchoring heavily operator-bound.

#Bridge reminder

If a representation shift matters in transport across contexts, note that the shift may alter what is preserved or salient. The bridge itself remains owned by F.9 and F.9.1.

#Review Matrix and Extension Tests

#Review matrix

A reviewer should ask:

• are all claimed factors visible in the publication or cited source;
• does any alias hide the factor bundle;
• is one factor being used as if it summarized the whole representation state;
• has representation talk started to replace articulation, closure, or anchoring claims.

#Local extension test

An additional factor is justified only if it captures a distinct representational question that cannot be reduced to locality, sparsity, or symbolicity. The extra factor should extend the bundle, not become a rival master axis.

#Migration test for legacy terminology

Legacy vocabularies often use "symbolic", "distributed", or "encoding basis" as if one term solved the whole classification problem. A conforming migration unpacks the term into explicit factor dockings and then checks whether any cross-facet claims were smuggled into the old label.

#Bundle-comparison reminder

Representation bundles may be compared across contexts only after the compared factors are explicit. If one context uses a compact local alias and another publishes the full factor bundle, require explicit docking before treating the two descriptions as commensurable.

#C.2.7:End