Inject a Turn, Not a Token: Learning Where to Redirect LLM Reasoning for Self-Correction

Tool errors or unavailable services cause your agent to abandon recoverable tasks.

What you would change

Inject repeatable tool failures and change only the recovery policy.

Inject repeatable tool failures and change only the recovery policy.

Compare against

Your current retry and fallback behavior under the same injected failures.

1. Locate the paper's implementation and check its license, model support, data needs, and hardware requirements. Stop here if you cannot obtain the required artifacts.
2. Save a small set of representative product tasks and expected outcomes. Keep evaluation tasks out of any training or tuning, and choose quality and cost limits before running.
3. Run the baseline and the proposed change on identical inputs. Keep other settings fixed and repeat paired runs to expose variation.
4. Save per-task outputs, failures, timings, and all usage costs. Review regressions before expanding the test.

What to measure

Recovered tasks, duplicate side effects, retries, time to recovery, and total cost.

When to keep it

Keep the change only if it meets your preselected quality and latency limits and improves the metric you care about after setup and operating costs. Otherwise retain the baseline. This is an exploration plan, not a validated recommendation.

recommendation

test

research area

failure recovery

technical change

Turn-level injection for LLM self-correction (UPTURN)

published

2026-10-09

The method requires no LLM fine-tuning and only a tiny trainable head, fitting constrained infrastructure, but its gains are demonstrated mainly on verifiable math reasoning, so the reader should validate it on their own workloads before production adoption.

1. claim 01Paperraft inference
   Read the brief interpretation

   Replaces LLM fine-tuning for self-correction (revision/feedback training) with a 130K-parameter MLP head on a frozen encoder that selects where a failed rollout went wrong, retains the prefix, injects an instruction as a new user turn, and regenerates the suffix. Costs are modest: training the small head requires failed-rollout data labeled only by final-answer correctness, and each correction consumes an additional partial regeneration, so inference budget grows with the number of interventions. It can fail when the MLP head's learned intervention points do not transfer to domains or reasoners unlike those in training, when the correctness signal is unavailable or noisy (limiting use outside verifiable tasks like math), and when extra regenerations erode the accuracy-per-compute advantage over simple majority voting.

   This is a Paperraft inference. It is not pinned to a verified PDF passage.

   Inspect the source
2. claim 02Paperraft inference

   Improves average accuracy over a 32-rollout baseline across four reasoners and six math benchmarks, outperforms self-correction and reflection-token injection baselines under matched correction budgets, and transfers to GPQA-Diamond without retraining; no multiplicative factor is reported.

   This is a Paperraft inference. It is not pinned to a verified PDF passage.

   Inspect the source