Reference library
Self-Refine: Iterative Refinement with Self-Feedback
Paper summary
Self-Refine uses the same language model to draft, critique, and revise outputs at inference time. Human and automatic evaluations across several tasks favor refined outputs without an additional parameter-training stage.
Relevance to the project
A comparator for transient self-feedback versus durable adaptation. Read alongside intrinsic-correction failures to identify which feedback and evaluation settings actually help.
Prompted refinement does not change weights or guarantee correctness.
Source record
- Reference ID
- arxiv-2303-17651
- Archived version
- arXiv preprint; retrieved version indicated in PDF
- Summary provenance
- reference-library/papers/arxiv-2303-17651/metadata.json · summary
- Review status
- initial orientation summary; not a critical full-paper review
- Reading scope
- Selected abstract/opening and available closing discussion/conclusion passages in the preserved text extract; broader project notes provide context. No reproduction of experiments, exhaustive methods review, or equation/figure audit.