LOVER creates an unsupervised logic-regularized verifier that reaches 95% of supervised verifier performance on reasoning tasks across 10 datasets.
Title resolution pending
5 Pith papers cite this work. Polarity classification is still indexing.
5
Pith papers citing it
representative citing papers
LLMs display high variance and major accuracy drops on GSM-Symbolic variants of grade-school math problems, indicating they replicate training patterns rather than execute logical reasoning.
Instructions trigger a production-centered mechanism in language models, with task-specific information stable in input tokens but varying strongly in output tokens and correlating with behavior.
COPSD improves mathematical reasoning in low-resource languages by having LLMs self-distill from their own high-resource English behavior via token-level divergence on rollouts with privileged crosslingual context.
TRACE aggregates answer consistency and confidence trajectory over multiple reasoning steps to decide when to halt inference, reducing token usage by 25-30% while keeping accuracy within 1-2% of full reasoning.
citing papers explorer
-
Logic-Regularized Verifier Elicits Reasoning from LLMs
LOVER creates an unsupervised logic-regularized verifier that reaches 95% of supervised verifier performance on reasoning tasks across 10 datasets.
-
GSM-Symbolic: Understanding the Limitations of Mathematical Reasoning in Large Language Models
LLMs display high variance and major accuracy drops on GSM-Symbolic variants of grade-school math problems, indicating they replicate training patterns rather than execute logical reasoning.
-
Instructions Shape Production of Language, not Processing
Instructions trigger a production-centered mechanism in language models, with task-specific information stable in input tokens but varying strongly in output tokens and correlating with behavior.
-
Crosslingual On-Policy Self-Distillation for Multilingual Reasoning
COPSD improves mathematical reasoning in low-resource languages by having LLMs self-distill from their own high-resource English behavior via token-level divergence on rollouts with privileged crosslingual context.
-
Efficient Test-Time Scaling via Temporal Reasoning Aggregation
TRACE aggregates answer consistency and confidence trajectory over multiple reasoning steps to decide when to halt inference, reducing token usage by 25-30% while keeping accuracy within 1-2% of full reasoning.