arXiv subjects

Junzhe Zhao

Publications and source records attributed to Junzhe Zhao.

5 recordsLinked to original sources

arXiv2026

Benchmarking Language Models for Statistical Problem Formulation

Large language models (LLMs) are increasingly used as assistants for statistical and data science work, yet existing evaluations largely assume the analysis target is already specified. In practice, users arrive with informal goals and heterogeneous data, leaving the model to decide what statistical task is implied and which data are relevant. We first formalize this upstream step as Statistical Problem Formulation and decompose it into two subtasks: (1) Statistical Problem Classification and (2) Variable Identification & Role Assignment. We then introduce StatFormBench, a benchmark built from five cross-domain statistics textbooks and a data science case library, covering diverse problem types, data representations, and scenario styles. It contains 1,013 samples spanning 20 coarse-grained and 85 fine-grained statistical problem categories. Across 14 open- and closed-source LLMs, the best zero-shot models reach only 72.0 fine-grained classification accuracy and 63.2 variable set overlap. No model performs consistently best across the two subtasks, while enhanced prompting strategies yield only limited or inconsistent gains. We release the benchmark data on Hugging Face at https://huggingface.co/datasets/THU-CongLab/StatFormBench and the evaluation code on GitHub at https://github.com/THU-CongLab/StatFormBench.

cs.AI↗

arXiv2025

CoDA: A Context-Decoupled Hierarchical Agent with Reinforcement Learning

Large Language Model (LLM) agents trained with reinforcement learning (RL) show great promise for solving complex, multi-step tasks. However, their performance is often crippled by "Context Explosion", where the accumulation of long text outputs overwhelms the model's context window and leads to reasoning failures. To address this, we introduce CoDA, a Context-Decoupled hierarchical Agent, a simple but effective reinforcement learning framework that decouples high-level planning from low-level execution. It employs a single, shared LLM backbone that learns to operate in two distinct, contextually isolated roles: a high-level Planner that decomposes tasks within a concise strategic context, and a low-level Executor that handles tool interactions in an ephemeral, isolated workspace. We train this unified agent end-to-end using PECO (Planner-Executor Co-Optimization), a reinforcement learning methodology that applies a trajectory-level reward to jointly optimize both roles, fostering seamless collaboration through context-dependent policy updates. Extensive experiments demonstrate that CoDA achieves significant performance improvements over state-of-the-art baselines on complex multi-hop question-answering benchmarks, and it exhibits strong robustness in long-context scenarios, maintaining stable performance while all other baselines suffer severe degradation, thus further validating the effectiveness of our hierarchical design in mitigating context overload.

cs.CL↗

arXiv2025

KARMA: Augmenting Embodied AI Agents with Long-and-short Term Memory Systems

Embodied AI agents responsible for executing interconnected, long-sequence household tasks often face difficulties with in-context memory, leading to inefficiencies and errors in task execution. To address this issue, we introduce KARMA, an innovative memory system that integrates long-term and short-term memory modules, enhancing large language models (LLMs) for planning in embodied agents through memory-augmented prompting. KARMA distinguishes between long-term and short-term memory, with long-term memory capturing comprehensive 3D scene graphs as representations of the environment, while short-term memory dynamically records changes in objects' positions and states. This dual-memory structure allows agents to retrieve relevant past scene experiences, thereby improving the accuracy and efficiency of task planning. Short-term memory employs strategies for effective and adaptive memory replacement, ensuring the retention of critical information while discarding less pertinent data. Compared to state-of-the-art embodied agents enhanced with memory, our memory-augmented embodied AI agent improves success rates by 1.3x and 2.3x in Composite Tasks and Complex Tasks within the AI2-THOR simulator, respectively, and enhances task execution efficiency by 3.4x and 62.7x. Furthermore, we demonstrate that KARMA's plug-and-play capability allows for seamless deployment on real-world robotic systems, such as mobile manipulation platforms.Through this plug-and-play memory system, KARMA significantly enhances the ability of embodied agents to generate coherent and contextually appropriate plans, making the execution of complex household tasks more efficient. The experimental videos from the work can be found at https://youtu.be/4BT7fnw9ehs. Our code is available at https://github.com/WZX0Swarm0Robotics/KARMA/tree/master.

cs.RO↗

arXiv2018

Kinetic Modelling and Inference of Hyperpolarized 13C Molecules in Cancer Metabolism

Hyperpolarized 13C-MRI allows real time observation of metabolism in vivo. Imaging sequences have been developed to follow the metabolism of [1-13C] pyruvate and extract reaction kinetics, which can show tumour treatment response. We applied the fitting model and algorithm for the imaging data of mice tumour models and determined error estimates for the parameters of interest. Data was least-squares fitted onto a two-site exchange model in MATLAB, followed by statistic computation to assess model performance. Inference through the application of MCMC was also performed. The modelling and inference process extracted quantitative information satisfactorily and reproducibly, demonstrating metabolic activity and intratumour heterogeneity. Finally, novel fitting methods were evaluated and further recommendations were made.

q-bio.BM↗

arXiv2018

Computational Modelling of Aquaporin Co-regulation in Cancer

A computational model of aquaporin regulation in cancer cells has been constructed as a Qualitative Network in the software BioModelAnalyzer (BMA). The model connects some important aquaporins expressed in human cancer to common phenotypes via a number of fundamental, dysregulated signalling pathways. Based on over 60 publications, this model can not only reproduce the results reported in a discrete, qualitative manner, but also reconcile the seemingly incompatible phenotype with research consensus by suggesting molecular mechanisms accountable for it. Novel predictions have also been made by mimicking real-life experiments in the model.

q-bio.MN↗