research-loop

An iterative research loop applies checks for novelty, baselines, datasets, and references to a proposal, then rewrites to address weaknesses. It presents changes with diffs and scores, repeating until the submission meets the target venue's standards.

Literature grounding in experimental design is the process of verifying claims against datasets and key papers. It ensures research proposals are well-supported by existing literature rather than relying on unsupported assertions during the refinement loop.

Provide an Obsidian research proposal file to the refinement loop to initiate the iteration cycle. The system reads the proposal, critiques it using a paper-critic model, grounds claims against datasets, and rewrites with revisions to strengthen the plan.

Yes, you can verify research baselines and datasets by running the proposal through a grounding and verification loop. The process checks claims against datasets and key papers to ensure all experimental baselines are properly validated before submission.

The best way to critique a research proposal for novelty and baselines is to apply an iterative, evidence-grounded loop. This approach uses a paper-critic model to evaluate the experimental design, then rewrites to address weaknesses and strengthen the plan.

Yes, the research loop outputs diffs and scores for manuscript revisions. After reading and critiquing the proposal, it presents changes with a diff, assigns scores, and provides a plan for another iteration to continuously strengthen the experimental plan.