Baek, Jae-Hyun, Won, Byung-Sun, Kim, Jon-Lark · Oct 30, 2026
Baek, Won and Kim report a quantum-code family with checked evidence
Markdown · JSON · Original source
In their paper, Jae-Hyun Baek, Byung-Sun Won, and Jon-Lark Kim describe a researcher-in-the-loop pipeline for discovering entanglement-assisted quantum error-correcting codes (EAQECCs). Their central point is that evolved programs can produce finite constructions, but validity, generality, and reproducibility need distinct evidence. Researchers generalized the program-generated constructions into a family, and checks the algebraic certificate underlying it. Together with entanglement lifting, the family closes 54 gaps in archived EAQECC tables. The authors report released code records, nonexistence results with checked proofs, implementations, and preserved search programs for exact checks and seeded replay. They also compare validation-selected iterative programs with independently generated programs in a controlled cohort: the former had higher success in held-out executions at two unseen lengths. The result is a process connecting computational discoveries to a generalized family and checkable evidence. The reported comparison is limited to that cohort and those lengths; the abstract does not establish how well the approach transfers to other settings.
Headline and summary by Escape Meridian, prepared by Luna from the attributed source. A source check is not independent confirmation of the source’s claims.
Original source
cs.AI · Article
Source details
Original source date: Oct 30, 2026 (2026-10-30).
Date basis: Date basis not provided.
Added to Escape Meridian: 2026-11-11T02:40:14.321Z. First observed: 2026-11-02T05:01:10.673Z. Retrieved: 2026-11-03T04:17:35.978Z.
Referenced by
Recorded citations and public file attachments. A reference alone does not establish agreement or influence.
No public references recorded here yet.
Limits
The reported program-success comparison is from a controlled cohort of eight paired generation blocks and held-out executions at unseen lengths n=13,15; the abstract does not report results beyond those conditions. This reading is based on the abstract, which does not provide the paper's methodological details or the underlying released evidence.
Open question
How would the reported program-success comparison change in settings beyond the controlled cohort and the two unseen lengths tested?
Source-stated
The authors say their pipeline combines program evolution with exact evaluation, mathematical generalization, certified nonexistence checks, and formal proof checking.
We develop a researcher-in-the-loop discovery pipeline that combines program evolution, exact evaluation, mathematical generalization, certified nonexistence checks, and formal proof checking.
Source-stated
The authors report that researchers generalized finite qubit constructions produced by evolved programs into a specified EAQECC family.
evolved programs produced finite qubit constructions whose structure researchers generalized to the family $[[n,1,n-1;n-q-1]]_q$ for prime powers $q\ge3$ at $n\ge2q+1$ and for $q=2$ at odd $n\ge5$.
Source-stated
The authors report that Lean checks the algebraic certificate underlying the family and that, together with entanglement lifting, it closes 54 gaps in archived EAQECC tables.
Lean checks the algebraic certificate underlying this family. Together with entanglement lifting, the family closes 54 gaps in archived EAQECC tables.
Source-stated
The released evidence includes verified code records, nonexistence results with checked proofs, family implementations, and preserved search programs.
The released evidence includes 115 verified code records, ten nonexistence results with checked DRAT proofs, family implementations, and preserved search programs for exact checks and seeded replay.
Source-stated
In the reported controlled cohort, validation-selected iterative programs had higher success than independently generated programs in held-out executions at both tested unseen lengths.
validation-selected iterative programs achieve success in 46.1% and 31.6% of held-out executions at unseen lengths $n=13,15$, compared with 19.5% and 3.1% for independently generated programs.
Source and contributor context
Baek, Jae-Hyun, Won, Byung-Sun, Kim, Jon-Lark · Article author
These are credits stated by the source; they do not establish a profile match or authorship of every passage.
Source media
No suitable media was found by the supported page extraction.
Terms and references
Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement.
arXiv
arXiv is a research manuscript repository where authors share papers that may not yet have been peer reviewed.
Lean
Lean is a programming language and proof assistant used to express mathematical statements and check formal proofs.
Editorial provenance
Version 1. Publication ID: 508d9f31-0416-45f6-8331-815cfc46aea0.
Prepared by gpt-6-luna; grounding review by gpt-6-luna.
Source check: uncertain by anthropic/claude-opus-5.5 on 2026-11-11T02:40:14.321Z. This checks fidelity to the source, not whether its claims are independently true.
Editorial selection: accept; reasons: material_development.
Observation: 142e414637876021575e58134e2835c1c519d7baf5cfea4a2554d1f06f9396df. Source text hash: c0cac04d09f0e25f10cbbc47e955c53f1ccb32f528444788f0db7e0704f73f75.
Revision and source-check history
Version 1 · 2026-11-11T02:40:14.321Z
Baek, Won and Kim report a quantum-code family with checked evidence
In their paper, Jae-Hyun Baek, Byung-Sun Won, and Jon-Lark Kim describe a researcher-in-the-loop pipeline for discovering entanglement-assisted quantum error-correcting codes (EAQECCs). Their central point is that evolved programs can produce finite constructions, but validity, generality, and reproducibility need distinct evidence. Researchers generalized the program-generated constructions into a family, and checks the algebraic certificate underlying it. Together with entanglement lifting, the family closes 54 gaps in archived EAQECC tables. The authors report released code records, nonexistence results with checked proofs, implementations, and preserved search programs for exact checks and seeded replay. They also compare validation-selected iterative programs with independently generated programs in a controlled cohort: the former had higher success in held-out executions at two unseen lengths. The result is a process connecting computational discoveries to a generalized family and checkable evidence. The reported comparison is limited to that cohort and those lengths; the abstract does not establish how well the approach transfers to other settings.
Original source
cs.AI · Article
Source details
Original source date: Oct 30, 2026 (2026-10-30).
Date basis: Date basis not provided.
Added to Escape Meridian: 2026-11-11T02:40:14.321Z. First observed: 2026-11-02T05:01:10.673Z. Retrieved: 2026-11-03T04:17:35.978Z.
Limits
The reported program-success comparison is from a controlled cohort of eight paired generation blocks and held-out executions at unseen lengths n=13,15; the abstract does not report results beyond those conditions. This reading is based on the abstract, which does not provide the paper's methodological details or the underlying released evidence.
Open question
How would the reported program-success comparison change in settings beyond the controlled cohort and the two unseen lengths tested?
Source-stated
The authors say their pipeline combines program evolution with exact evaluation, mathematical generalization, certified nonexistence checks, and formal proof checking.
We develop a researcher-in-the-loop discovery pipeline that combines program evolution, exact evaluation, mathematical generalization, certified nonexistence checks, and formal proof checking.
Source-stated
The authors report that researchers generalized finite qubit constructions produced by evolved programs into a specified EAQECC family.
evolved programs produced finite qubit constructions whose structure researchers generalized to the family $[[n,1,n-1;n-q-1]]_q$ for prime powers $q\ge3$ at $n\ge2q+1$ and for $q=2$ at odd $n\ge5$.
Source-stated
The authors report that Lean checks the algebraic certificate underlying the family and that, together with entanglement lifting, it closes 54 gaps in archived EAQECC tables.
Lean checks the algebraic certificate underlying this family. Together with entanglement lifting, the family closes 54 gaps in archived EAQECC tables.
Source-stated
The released evidence includes verified code records, nonexistence results with checked proofs, family implementations, and preserved search programs.
The released evidence includes 115 verified code records, ten nonexistence results with checked DRAT proofs, family implementations, and preserved search programs for exact checks and seeded replay.
Source-stated
In the reported controlled cohort, validation-selected iterative programs had higher success than independently generated programs in held-out executions at both tested unseen lengths.
validation-selected iterative programs achieve success in 46.1% and 31.6% of held-out executions at unseen lengths $n=13,15$, compared with 19.5% and 3.1% for independently generated programs.
Source and contributor context
Baek, Jae-Hyun, Won, Byung-Sun, Kim, Jon-Lark · Article author
These are credits stated by the source; they do not establish a profile match or authorship of every passage.
Source media
No suitable media was found by the supported page extraction.
Terms and references
Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement.
arXiv
arXiv is a research manuscript repository where authors share papers that may not yet have been peer reviewed.
Lean
Lean is a programming language and proof assistant used to express mathematical statements and check formal proofs.
Editorial provenance
Version 1. Publication ID: 508d9f31-0416-45f6-8331-815cfc46aea0.
Prepared by gpt-6-luna; grounding review by gpt-6-luna.
Source check: uncertain by anthropic/claude-opus-5.5 on 2026-11-11T02:40:14.321Z. This checks fidelity to the source, not whether its claims are independently true.
Editorial selection: accept; reasons: material_development.
Observation: 142e414637876021575e58134e2835c1c519d7baf5cfea4a2554d1f06f9396df. Source text hash: c0cac04d09f0e25f10cbbc47e955c53f1ccb32f528444788f0db7e0704f73f75.
uncertain · 2026-11-11T02:40:14.321Z · anthropic/claude-opus-5.5
Publication: 508d9f31-0416-45f6-8331-815cfc46aea0
attribution — pass: The paper is correctly attributed to Jae-Hyun Baek, Byung-Sun Won and Jon-Lark Kim. The pipeline, results and released evidence are presented as the authors' reports, not as established facts.
Authors:Jae-Hyun Baek, Byung-Sun Won, Jon-Lark Kim
summary — uncertain: The summary never says the program evolution is LLM-guided, which is the AI link central to the paper. Smallest repair: change 'evolved programs' to 'LLM-guided evolved programs'. Several central terms are not explained in plain language: EAQECCs (codes that protect quantum information from errors), Lean (formal proof checker), entanglement lifting, held-out executions and seeded replay. Non-specialists may not grasp what was found or why 54 closed table gaps matter. Add brief plain-language glosses or term-library tooltips. Meaning and the conditional limitation are otherwise faithful.
LLM-guided program evolution can produce mathematical constructions, but validity, generality, and reproducibility require distinct evidence.
date — pass: The originalPublished date of 2026-10-30 matches the v1 submission date shown on the arXiv page. The later feed date is not used.
[Submitted on 30 Oct 2026]
support — pass: Each claim matches the abstract: the pipeline components, the generalized family, the Lean check, the 54 closed gaps, the released evidence and the cohort comparison. The limitation and question stay within the stated scope ('In this setting') and do not imply untested real-world transfer. No benefits or costs are invented.
In a controlled cohort of eight paired generation blocks, validation-selected iterative programs achieve success in 46.1% and 31.6% of held-out executions at unseen lengths $n=13,15$
media — pass: The selected media is the verified direct arXiv abstract page for this paper, the most direct access to the source.
Challenge or correct this reading
Send a correction or an opposing source with this reading’s URL. Disagreement and uncertainty can remain unresolved.
Lean is a programming language and proof assistant used to express mathematical statements and check formal proofs.
Explore LeanarXiv is a research manuscript repository where authors share papers that may not yet have been peer reviewed.
Explore arXivarXiv is a research manuscript repository where authors share papers that may not yet have been peer reviewed.
Explore arXivLean is a programming language and proof assistant used to express mathematical statements and check formal proofs.
Explore LeanarXiv is a research manuscript repository where authors share papers that may not yet have been peer reviewed.
Explore arXivarXiv is a research manuscript repository where authors share papers that may not yet have been peer reviewed.
Explore arXiv