analiza aiqbench
ReTreVal introduces a training-free framework that enables fixed LLMs to leverage cross-problem learning by retaining and revising error contexts during inference, effectively allowing smaller models to compete with larger single-pass systems. By utilizing adaptive tree exploration and self-rewriting memory, the method achieves significant performance gains on MATH-500 and MMLU-Pro without requiring gradient updates. This approach addresses a critical inefficiency in current reasoning frameworks where failure data is typically discarded between problems.
ReTreVal: Reasoning Tree with Validation and Cross-Problem Memory for Large Language Models
arXiv:2601.02880v2 Announce Type: replace Abstract: Every existing inference-time reasoning framework discards all failure context at problem boundaries, leaving a model solving problem 500 no wiser than it was on problem 1. We present ReTreVal (Reasoning Tree with Validation), a training-free framework that closes this gap through adaptive tree exploration with tool-augmented node refinement, typed-failure backtracking that injects categorized error context into the recovered branch, and a self-rewriting memory that accumulates and revises strategy entries across problems, enabling inference-time cross-problem learning on any fixed, unmodified LLM without fine-tuning. ReTreVal achieves 85.8% pass@1 on MATH-500 (+8.6 pp over Zero-Shot CoT, +8.6 pp over the strongest baseline Self-Refine) and 54.4% on MMLU-Pro (+15.3 pp over Self-Refine), with a 3.4:1 win-to-regression ratio confirming genuine error recovery rather than noise. These capabilities, previously requiring gradient updates, allow a 32B model to compete with much larger single-pass systems.
Abstract:Every existing inference-time reasoning framework discards all failure context at problem boundaries, leaving a model solving problem 500 no wiser than it was on problem 1. We present ReTreVal (Reasoning Tree with Validation), a training-free framework that closes this gap through adaptive tree exploration with tool-augmented node refinement, typed-failure backtracking that injects categorized error context into the recovered branch, and a self-rewriting memory that accumulates and revises strategy entries across problems, enabling inference-time cross-problem learning on any fixed, unmodified LLM without fine-tuning. ReTreVal achieves 85.8% pass@1 on MATH-500 (+8.6 pp over Zero-Shot CoT, +8.6 pp over the strongest baseline Self-Refine) and 54.4% on MMLU-Pro (+15.3 pp over Self-Refine), with a 3.4:1 win-to-regression ratio confirming genuine error recovery rather than noise. These capabilities, previously requiring gradient updates, allow a 32B model to compete with much larger single-pass systems.
Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
arXiv:2601.02880 [cs.AI]
(or arXiv:2601.02880v2 [cs.AI] for this version)
https://doi.org/10.48550/arXiv.2601.02880 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 233 cuvinte
Subiecte conexe
ReTreVal: Reasoning Tree with Validation and Cross-Problem Memory for Large Language Models
arXiv · 9 iun. 2026
Performance Profiling on AMD GPUs – Part 5: Profiling-Driven Kernel Optimization with an AI Code-Assist Tool
AMD · 16 iul. 2026
Presentation: From Hype to Strong Foundations: What the Rise, Fall and Resurgence of Agents Can Teach Us About Outlasting the Cycle
InfoQ · 17 iun. 2026
News
Quantum Computing Report · 25 iul. 2026