pllmPublications and claims
Learn which evidence and review a PLLM research claim needs before publication.
PLLM distinguishes a reproduction, an engineering improvement, and a scientific contribution. Reproducing prior work, cleaning code, moving code to Rust, reporting an unmatched speed result, weakening privacy or quality, or hiding system costs is not by itself a new scientific contribution.
A proposed new contribution needs a distinct method, a documented search of prior work, a clear novelty argument, completed review gates, and a matched improvement without hidden regressions. A paper claim also needs comparable evidence, scoped assurance, negative results, and explicit limitations. Each claim links to its source, PLLM implementation, plan, cohort, evidence, and document build.
The PLLM Research Paper and PLLM Whitepaper are scoped historical records. Future claims still require the same evidence and review.