paper-evaluate

To evaluate a robotics paper, apply the Appendix B rubric (B.1-B.7) to the full text or metadata. This generates structured scores with supporting evidence, task chain extraction, and honesty flags for automated decision making.

Task chain extraction maps a paper's logical flow from SIGNIFICANCE through TASK, PROBLEM DEFINITION, CHALLENGE, APPROACH, to SCOPE. This structured assessment populates evaluation records for systematic literature surveys.

Yes, you can run looped evaluations across multiple robotics papers to populate an evaluations.jsonl file. This supports systematic literature surveys by generating canonical per-paper assessments with evidence and optional human flags.

You need access to either the full text or metadata of the robotics paper to perform the evaluation. The rubric analyzes this input to produce significance assessments and honesty flags across the B.1-B.7 criteria.

Honesty flags are markers generated during the rubric evaluation to indicate potential issues or inconsistencies in a robotics paper. They accompany the B.1-B.7 scores and evidence to support automated decision making and literature review accuracy.

You need rubric-based paper evaluation when conducting systematic literature surveys or automated decision making that requires canonical, structured assessments. It ensures consistent scoring across B.1-B.7 criteria with extracted evidence rather than subjective manual review.