Search papers, labs, and topics across Lattice.
This paper introduces PaperJury, a closed-loop system designed for the adversarial review of LaTeX computer science papers, which addresses the limitations of existing writing assistants by implementing a deterministic orchestration for managing critiques and revisions. The system features a unique combination of bounded holistic review and contestability-based routing, ensuring that authors can navigate through valid and invalid concerns effectively. In a comparative evaluation against four baselines, PaperJury demonstrated superior issue quality and edit safety, reinforcing the importance of deterministic processes in the review workflow.
PaperJury reveals that a deterministic orchestration can significantly enhance the safety and effectiveness of the academic paper review process.
Pre-submission hardening of human-authored LaTeX computer science papers differs from drafting assistance because it requires adversarial whole-paper review, explicit no-fix outcomes, and bounded artifact-safe revision. Existing writing assistants, critique generators, and judge-centered loops lack durable issue identity across rounds, deterministic routing from critique to adjudication, and manuscript control that can reject invalid concerns or defer author-dependent ones. We present PaperJury, a closed-loop review-verdict-revise-verify system built on a deterministic-versus-semantic split: deterministic orchestration manages decomposition, a frozen claim spine, a durable ledger, routing, stopping, and exact-once patch application, while semantic agents are limited to bounded review, judgment, and repair. PaperJury combines bounded holistic review, contestability-based routing, a due-process trial, and risk-proportional guard chains for anchor-bounded edits, yielding terminal outcomes of invalid-drop, valid-fixable, and author-required. In a two-arm expert-review evaluation on held-out Vision, natural language processing, and machine learning papers against four baselines, we assess issue quality, verdict and routing quality, edit safety, convergence behavior, and cost, supporting the thesis that load-bearing safety and completion logic should reside in deterministic orchestration rather than model discretion. PaperJury is available at https://github.com/u7079256/paperjury.
