Recent advances in block diffusion language models have demonstrated competitive performance and strong scalability on reasoning tasks. However, their test-time compute allocation remains largely unexplored, leaving a critical speed-effectiveness trade-off unresolved in long Chain-of-Thought reasoning. To address this, we propose a unified test-time compute allocation framework that introduces adaptivity in both step-wise decoding and blockwise generation. At the decoding level, we propose Bounded Adaptive Confidence Decoding (BACD), a difficulty-aware sampling strategy that dynamically adjusts denoising based on model confidence, accelerating inference while controlling error accumulation. Beyond step-wise adaptivity, we introduce the Think Coarse, Critic Fine (TCCF) paradigm that allocates large block sizes for exploratory thinking and smaller block sizes for precise refinement. To stabilize training under varying block configurations, we adopt Progressive Block Size Extension, which mitigates quality degradation when scaling up block sizes. Extensive evaluations on six benchmarks show that our TDAR-8B model with BACD and TCCF achieves a 2.38$\times$ speedup and +3.4% average accuracy over the strong TraDo-8B baseline, unlocking the potential of block diffusion in complex reasoning.
DRQA: Dynamic Reasoning Quota Allocation for Controlling Overthinking in Reasoning Large Language Models
The Hong Kong University of Science and Technology (Guangzhou)(香港科学与技术大学(广州));The Hong Kong University of Science and Technology(香港科学与技术大学);Tsinghua University(清华大学);Fudan University(复旦大学);Renmin University of China(中国人民大学);Shanghai Jiao Tong University(上海交通大学)
SelfBudgeter: Adaptive Token Allocation for Efficient LLM Reasoning
State Key Laboratory for Multimedia Information Processing, School of Computer Science, Peking University(多媒体信息处理国家重点实验室,计算机学院,北京大学);BandAI, Bytedance(字节跳动BandAI)
CVPD at QIAS 2026: RAG-Guided LLM Reasoning for Al-Mawarith Share Computation and Heir Allocation
ETIS (UMR 8051), CY Cergy Paris Univ., ENSEA, CNRS(ETIS(UMR 8051),CY塞尔吉巴黎大学,ENSEA,法国国家科学研究中心);VSC Laboratory, Department of Electronics and Automation, University of Biskra(VSC实验室,电子与自动化系,比斯克拉大学);Computer Engineering Dept, Sana'a Community College(计算机工程系,萨那社区学院);CNR-ISASI "E. Caianiello"(意大利国家研究委员会-智能系统自动化研究所(ISASI)"E. Caianiello");Ikerbasque(伊克尔巴斯克)
BESA: Pruning Large Language Models with Blockwise Parameter-Efficient Sparsity Allocation
The University of Hong Kong(香港大学);OpenGVLab, Shanghai AI Laboratory(上海人工智能实验室开放智能研究院);Simon Fraser University(西蒙菲莎大学);Southern University of Science and Technology(南方科技大学)
Unlocking the Working Memory of Large Language Models for Latent Reasoning
ELLIS Unit Linz and LIT AI Lab, Institute for Machine Learning, Johannes Kepler University Linz, Austria(林茨ELLIS单元和LIT AI实验室,机器学习研究所,约翰·凯撒大学林茨,奥地利);NXAI GmbH, Linz, Austria(NXAI公司,林茨,奥地利)
Incentivizing Dual Process Thinking for Efficient Large Language Model Reasoning
Gaoling School of Artificial Intelligence, Renmin University of China(中国人民大学人工智能学院);Department of Computer Science, National University of Singapore(新加坡国立大学计算机科学系);Ant Group(蚂蚁集团)