Execution Sufficiency and State Sufficiency: Record Sufficiency in Reflexive Research Systems
Description
AI-mediated research records often preserve enough information to replay a process without preserving enough information to resume it. This article formalizes that distinction through execution sufficiency and state sufficiency. A record is execution-sufficient when it determines a replayable continuation class; it is state-sufficient when it determines the operative state required for continuation. Using a compact model of operative state, record projection, replay, and restoration, the article proves that replayability and resumability separate whenever the record projection collapses distinct operative states into the same externalized trace. It then shows constructively that augmenting the artifact family can remove the relevant collapse on a targeted state class. The architectural consequence is direct: future AI-mediated research systems require state-bearing records, not merely replayable logs.
Notes
Keywords
reflexive research systems
record sufficiency
execution sufficiency
state sufficiency
replayability
resumability
AI-mediated research
Access right
Open Access
License
CC BY 4.0
(only use this if that is your intended license)
DOI fields
Concept DOI
10.5281/zenodo.18902418
Version DOI
10.5281/zenodo.19042930
Recommended “Related identifiers” entry
If Zenodo gives you a manual related-identifier field, use:
-
Identifier:
10.5281/zenodo.18902418 -
Relation:
isVersionOf -
Resource type:
publication-article
If Zenodo already auto-links the concept DOI for the versioned record, you do not need to add this manually.
Recommended additional description note
You can append this to the description if you want the archive record to better explain the package contents:
This Zenodo record includes the article PDF, manuscript source, Graphviz/DOT source files for Figures A10-01 to A10-03, rendered figure assets, and release-packaging materials. The source transcript corpus and figure-generation artifacts function as design-log evidence and reproducible diagram assets for the manuscript.
Optional fields I recommend filling
Subjects / communities
-
Applied Mathematics
-
Mathematical Modeling
-
AI and Society
-
Research Methods
-
Computational Research Workflows
Notes
-
Funding: none
-
Competing interests: none
-
Generative AI tools were used during interactive manuscript development and editorial refinement; all formal claims, theorem structure, figure logic, and final manuscript decisions were reviewed and curated by the author.
Files
2026-03-15-P10 SUFFICIENCY MODELS.pdf
Additional details
Dates
- Created
-
2026-03-15