Record · analyzed
Large Language Models Cannot Self-Correct Reasoning Yet
Without outside feedback, models asked to correct their own reasoning often get worse rather than better.
1
Listed
Added to the list. How entries were checked when they were listed is in the journal.
2
Collected
| File | Source | SHA-256 |
|---|---|---|
| paper PDF2310.01798.pdf | arxiv.org/pdf/2310.01798 | 15e1731e255ec6b217792a1e077926807da77e8e46b69e6c57dda30d67794761 |
To check a file, download it from its source and compare its SHA-256 with the one above. Collected copies are kept in the editor's RSI system; manifest: design-life/literature/rsi-workflow-improvement/sources.json.
3
Analyzed
Quoted verbatim from design-life/literature/rsi-workflow-improvement/comparison.md (2026-10-07). Parts about applying the paper to the editor's private work are left out.
From “Evidence for improving reusable agent workflows” (2026-10-07)
Inspected edition: Large Language Models Cannot Self-Correct Reasoning Yet, v2, 14 March 2024; PDF pp. 2–6
Mechanism and reported evidence: Removes privileged correctness feedback and checks alternative prompts and call budgets. Table 3: GPT-4 GSM8K accuracy falls from 95.5% with one call to 91.5% after a correction round with three calls, then 89.0% with five calls.
4
Replicated
Not done yet. Not attempted yet.
5
Audited
Not done yet. Not attempted yet.
The editor's view
None recorded yet.