RubricRefine is a training-free pre-execution method that creates rubrics to score and fix inter-tool contract violations in agent code, reaching 0.86 average on M3ToolEval across seven models with zero executions and lower latency.
When can LLMs actually correct their own mistakes? A critical survey of self-correction of LLMs
8 Pith papers cite this work. Polarity classification is still indexing.
8
Pith papers citing it
years
2026 8representative citing papers
Many-shot ICL with LLMs matches or exceeds supervised BERT on NER and generates high-quality labels for low-resource settings, producing ~10% absolute F1 gains when used to fine-tune BERT.
LORIS detects local reasoning errors in LLM-generated proofs for loop invariants by translating natural-language steps to first-order logic implications and using invalid implications to refine the invariants, achieving 93.1% success on 460 C programs.
Weighted rules extend stable model semantics to support probabilistic reasoning, model ranking, and statistical inference in answer set programs.
ReflectCAP distills model-specific hallucination and oversight patterns into Structured Reflection Notes that steer LVLMs toward more factual and complete image captions, reaching the Pareto frontier on factuality-coverage trade-offs.
A two-rate measurement (correction c and corruption γ) for LLM protocol steps predicts accuracy changes from paired correctness bits and flags three failure modes including mixture shift on GSM8K.
ReBias-Lens shows LLM self-reflection produces layer-wise smoothing of global valence fluctuations that reduces behavioral bias overall, yet selectively locks in and amplifies certain category-specific biases.
citing papers explorer
-
Scaling Performance and Low-Resource Annotation with Many-Shot In-Context Learning for Named Entity Recognition
Many-shot ICL with LLMs matches or exceeds supervised BERT on NER and generates high-quality labels for low-resource settings, producing ~10% absolute F1 gains when used to fine-tune BERT.