analiza aiqbench
ThinkBooster addresses the fragmentation in test-time compute (TTC) scaling by offering a unified Python library, a joint performance-efficiency benchmark, and an OpenAI-compatible proxy for drop-in integration. The framework provides practical gains on mathematical and coding tasks while enabling visual debugging of reasoning trajectories to better analyze quality-cost trade-offs. This release is significant for engineers seeking standardized tools to optimize inference-time compute allocation in production LLM applications.
ThinkBooster: A Unified Framework for Seamless Test-Time Scaling of LLM Reasoning
arXiv:2606.06915v1 Announce Type: cross Abstract: Test-time compute (TTC) scaling has emerged as a powerful paradigm for improving large language model (LLM) reasoning by allocating additional compute during inference, e.g., via multi-sample generation and verifier-based reranking. Existing TTC scaling strategies and reasoning scorers remain fragmented, evaluated under inconsistent protocols, and are rarely analyzed through the lens of quality-cost trade-offs. We introduce ThinkBooster, a unified framework for seamless test-time compute scaling of LLM reasoning, which consists of (i) a modular Python library implementing state-of-the-art TTC scaling strategy and scorer families, (ii) a benchmark that jointly evaluates performance and computational efficiency, and (iii) a deployable OpenAI-compatible proxy service that enables drop-in integration of adaptive reasoning into real-world applications. We further provide a demo visual debugger for inspecting the reasoning trajectories, intermediate selection decisions, and alternative reasoning paths. Empirical results on mathematical and coding tasks reveal the performance-compute trade-offs of TTC scaling strategies and scoring methods and demonstrate that ThinkBooster provides practical gains in real-world tasks. The code is available online under an MIT license.
Abstract:Test-time compute (TTC) scaling has emerged as a powerful paradigm for improving large language model (LLM) reasoning by allocating additional compute during inference, e.g., via multi-sample generation and verifier-based reranking. Existing TTC scaling strategies and reasoning scorers remain fragmented, evaluated under inconsistent protocols, and are rarely analyzed through the lens of quality-cost trade-offs. We introduce ThinkBooster, a unified framework for seamless test-time compute scaling of LLM reasoning, which consists of (i) a modular Python library implementing state-of-the-art TTC scaling strategy and scorer families, (ii) a benchmark that jointly evaluates performance and computational efficiency, and (iii) a deployable OpenAI-compatible proxy service that enables drop-in integration of adaptive reasoning into real-world applications. We further provide a demo visual debugger for inspecting the reasoning trajectories, intermediate selection decisions, and alternative reasoning paths. Empirical results on mathematical and coding tasks reveal the performance-compute trade-offs of TTC scaling strategies and scoring methods and demonstrate that ThinkBooster provides practical gains in real-world tasks. The code is available online under an MIT license.
Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
arXiv:2606.06915 [cs.CL]
(or arXiv:2606.06915v1 [cs.CL] for this version)
https://doi.org/10.48550/arXiv.2606.06915 arXiv-issued DOI via DataCite (pending registration)
From: Vladislav Smirnov Mr. [view email] [v1] Fri, 5 Jun 2026 05:28:46 UTC (1,180 KB)
Text integral sub licența deschisă a publicației · 206 cuvinte