arXiv:2504.00613v2 Announce Type: replace Abstract: Finding deletion-correcting codes of maximum size has been an open problem for over 70 years, even for a single deletion. We adapt FunSearch, a large language model (LLM)-guided evolutionary search, to discover functions that construct deletion-correcting codes at short code lengths. For a single deletion, our search finds a function that we prove constructs the conjectured-optimal Varshamov-Tenengolts code. For multiple deletions and quaternary edit codes, the discovered functions improve on prior explicit, search-based, and neural construct
The increasing sophistication of large language models is enabling their application to complex combinatorial search problems that were previously intractable, such as finding optimal codes.
This development demonstrates a tangible advancement in using AI for fundamental scientific and engineering problems, specifically in error correction where robust data transmission is critical.
The use of LLM-guided evolutionary search can significantly accelerate the discovery of optimal or near-optimal solutions for long-standing combinatorial problems, potentially leading to more efficient coding schemes.
- · AI/ML Research Institutions
- · Data Transmission Technologies
- · Information Theory Researchers
- · Telecommunications Industry
- · Traditional Brute-Force Algorithm Approaches
- · Manual Code Design
The immediate effect is a new method for generating highly efficient deletion-correcting codes, improving data integrity.
This methodology could be generalized to solve other hard combinatorial optimization problems across various scientific and engineering domains.
Improved error correction could enable new forms of high-reliability, low-latency communication systems or data storage paradigms.
This signal links to a primary source. Continuum Brief monitors and indexes it as part of the live intelligence stream — we do not republish source content.
Read at arXiv cs.AI