Zhipeng Ma

Dingwei Chen, Zefang Zong, Zhipeng Ma et al.

Agentic reinforcement learning (RL) has proven effective for training LLM-based agents with external tool-use capabilities. However, we identify that agentic RL training induces increasing redundant tool calls and blurs the model's intrinsic knowledge boundary, where the model fails to distinguish when tools are needed versus when parametric knowledge suffices. Existing solutions based on reward shaping create coarse-grained optimization targets that tend to incentivize indiscriminate tool-call suppression, leading to reward hacking. In this paper, we propose AKBE (Agentic Knowledge Boundary Enhancement), an on-policy method that dynamically probes the model's intrinsic knowledge boundary through dual-path (with-tool and no-tool) rollouts during training. We define the knowledge boundary as the per-instance determination of whether tools are required and the minimum tool calls necessary. By comparing correctness across paths, AKBE categorizes trajectories and constructs targeted supervisory signals that guide efficient tool-use patterns for each question. These signals are integrated seamlessly into the agentic RL training loop. Experiments on seven QA benchmarks demonstrate that AKBE improves task accuracy by +1.85 on average and reduces tool calls by 18% over standard agentic RL, yielding 25% higher tool productivity without any accuracy-efficiency trade-off. Further analysis suggests its plug-and-play compatibility across different RL algorithms and the mechanism of each signal category. Our code is available at https://github.com/CuSO4-Chen/AKBE.

Leo Luo, Haining Xie, Siqi Shen et al.

Text-to-SQL on complex schemas is unreliable on a single pass, so recent systems generate multiple SQL candidates and let voting filter out errors. Yet voting alone is not enough, because the multi-candidate recipe has three coupled weaknesses: 1) sampling more from a single generator produces increasingly redundant candidates, 2) existing pipelines apply one generic correction to every non-clean execution result, while runtime errors, timeouts, and empty results each indicate a different distance from correctness, and 3) existing selectors rely on a single angle such as result-majority voting or pairwise SQL comparison, missing what other angles would have caught. We present SIRIUS-SQL, which addresses all three weaknesses. A difficulty-smoothing RL recipe trains SIRIUS-32B to generate diverse executable SQL candidates, paired with a generalist LLM that fills in gaps left by the specialist. An execution-grounded lifecycle classifies each outcome and applies targeted repair before candidates re-enter the pool. A confidence-gated hybrid selector combines execution-result agreement with pairwise SQL-form judgment, escalating only near-tied cases to a deterministic structural check. SIRIUS-SQL reaches 75.88% on BIRD dev and 91.20% on SPIDER test. Two of three generalist pairings surpass Agentar-Scale-SQL, the strongest published multi-candidate system on BIRD dev.

Zhipeng Ma, Zheyan Tu, Xinhai Chen et al. · baidu, tsinghua

Trajectory representation learning plays a pivotal role in supporting various downstream tasks. Traditional methods in order to filter the noise in GPS trajectories tend to focus on routing-based methods used to simplify the trajectories. However, this approach ignores the motion details contained in the GPS data, limiting the representation capability of trajectory representation learning. To fill this gap, we propose a novel representation learning framework that Joint GPS and Route Modelling based on self-supervised technology, namely JGRM. We consider GPS trajectory and route as the two modes of a single movement observation and fuse information through inter-modal information interaction. Specifically, we develop two encoders, each tailored to capture representations of route and GPS trajectories respectively. The representations from the two modalities are fed into a shared transformer for inter-modal information interaction. Eventually, we design three self-supervised tasks to train the model. We validate the effectiveness of the proposed method on two real datasets based on extensive experiments. The experimental results demonstrate that JGRM outperforms existing methods in both road segment representation and trajectory representation tasks. Our source code is available at Anonymous Github.

Runchuan Zhu, Zinco Jiang, Jiang Wu et al.

Refusal-Aware Instruction Tuning (RAIT) aims to enhance Large Language Models (LLMs) by improving their ability to refuse responses to questions beyond their knowledge, thereby reducing hallucinations and improving reliability. Effective RAIT must address two key challenges: firstly, effectively reject unknown questions to minimize hallucinations; secondly, avoid over-refusal to ensure questions that can be correctly answered are not rejected, thereby maintain the helpfulness of LLM outputs. In this paper, we address the two challenges by deriving insightful observations from the gradient-based perspective, and proposing the Gradient-driven Refusal Aware Instruction Tuning Framework GRAIT: (1) employs gradient-driven sample selection to effectively minimize hallucinations and (2) introduces an adaptive weighting mechanism during fine-tuning to reduce the risk of over-refusal, achieving the balance between accurate refusals and maintaining useful responses. Experimental evaluations on open-ended and multiple-choice question answering tasks demonstrate that GRAIT significantly outperforms existing RAIT methods in the overall performance. The source code and data will be available at https://github.com/opendatalab/GRAIT .

Dingwei Chen, Zefang Zong, Zhipeng Ma et al.

Reinforcement learning for agentic large language models (LLMs) typically relies on a sparse, trajectory-level outcome reward, making it difficult to evaluate the contribution of individual tool-calls within multi-turn interactions. Existing approaches to such process credit assignment either depend on separate external process reward models that introduce additional consumption, or tree-based structural rollout that merely redistributes the outcome signal while constraining trajectory diversity. A promising alternative leverages the per-turn change in the policy's predicted probability of the ground-truth, termed Information Gain (IG), as an intrinsic process signal without an external evaluator. However, prior work on leveraging IG signals within the RL training loop faces three systematic challenges: normalizing across turns that face heterogeneous positional contexts can distort the relative standing of individual turns, accumulating a variable number of terms causes advantage magnitudes to drift with trajectory depth, and a fixed clipping range governs policy updates identically for turns with vastly different IG signals. In this paper, we propose A$^2$TGPO (Agentic Turn-Group Policy Optimization with Adaptive Turn-level Clipping), which retains IG as the intrinsic signal but re-designs how it is normalized, accumulated, and consumed: (i) turn-group normalization: normalizes IG within each (prompt, turn-index) group so that each turn is compared only against peers at the same interaction depth; (ii) variance-rescaled discounted accumulation: divides cumulative normalized IG by square root of accumulated terms to keep advantage magnitudes comparable across turn positions; and (iii) adaptive turn-level clipping: modulates each turn's clipping range based on its normalized IG, widening the update region for informative turns and narrowing it for uninformative ones.

Zhipeng Ma, Bo Nørregaard Jørgensen, Zheng Ma

The transportation sector remains a major contributor to greenhouse gas emissions. The understanding of energy-efficient driving behaviors and utilization of energy-efficient driving strategies are essential to reduce vehicles' fuel consumption. However, there is no comprehensive investigation into energy-efficient driving behaviors and strategies. Furthermore, many state-of-the-art AI models have been applied for the analysis of eco-friendly driving styles, but no overview is available. To fill the gap, this paper conducts a thorough literature review on ecological driving behaviors and styles and analyzes the driving factors influencing energy consumption and state-of-the-art methodologies. With a thorough scoping review process, the methodological and related data are compared. The results show that the factors that impact driving behaviors can be summarized into eleven features including speed, acceleration, deceleration, pedal, and so on. This paper finds that supervised/unsupervised learning algorithms and reinforcement learning frameworks have been popularly used to model the vehicle's energy consumption with multi-dimensional data. Furthermore, the literature shows that the driving data are collected from either simulators or real-world experiments, and the real-world data are mainly stored and transmitted by meters, controller area networks, onboard data services, smartphones, and additional sensors installed in the vehicle. Based on driving behavior factors, driver characteristics, and safety rules, this paper recommends nine energy-efficient driving styles including four guidelines for the drivers' selection and adjustment of the vehicle parameters, three recommendations for the energy-efficient driving styles in different driving scenarios, and two subjective suggestions for different types of drivers and employers.

Zhipeng Ma, Marco Kemmerling, Daniel Buschmann et al.

Causal inference is a fundamental research topic for discovering the cause-effect relationships in many disciplines. However, not all algorithms are equally well-suited for a given dataset. For instance, some approaches may only be able to identify linear relationships, while others are applicable for non-linearities. Algorithms further vary in their sensitivity to noise and their ability to infer causal information from coupled vs. non-coupled time series. Therefore, different algorithms often generate different causal relationships for the same input. To achieve a more robust causal inference result, this publication proposes a novel data-driven two-phase multi-split causal ensemble model to combine the strengths of different causality base algorithms. In comparison to existing approaches, the proposed ensemble method reduces the influence of noise through a data partitioning scheme in the first phase. To achieve this, the data are initially divided into several partitions and the base algorithms are applied to each partition. Subsequently, Gaussian mixture models are used to identify the causal relationships derived from the different partitions that are likely to be valid. In the second phase, the identified relationships from each base algorithm are then merged based on three combination rules. The proposed ensemble approach is evaluated using multiple metrics, among them a newly developed evaluation index for causal ensemble approaches. We perform experiments using three synthetic datasets with different volumes and complexity, which are specifically designed to test causality detection methods under different circumstances while knowing the ground truth causal relationships. In these experiments, our causality ensemble outperforms each of its base algorithms. In practical applications, the use of the proposed method could hence lead to more robust and reliable causality results.

Zhipeng Ma, Bo Nørregaard Jørgensen, Zheng Ma

Public transportation is a major source of greenhouse gas emissions, highlighting the need to improve bus fuel efficiency. Clustering algorithms assist in analyzing fuel efficiency by grouping data into clusters, but irrelevant features may complicate the analysis and choosing the optimal number of clusters remains a challenging task. Therefore, this paper employs the Gaussian mixture models to cluster the solo fuel-efficiency dataset. Moreover, an integration method that combines the Silhouette index, Calinski-Harabasz index, and Davies-Bouldin index is developed to select the optimal cluster numbers. A dataset with 4006 bus trips in North Jutland, Denmark is utilized as the case study. Trips are first split into three groups, then one group is divided further, resulting in four categories: extreme, normal, low, and extremely low fuel efficiency. A preliminary study using visualization analysis is conducted to investigate how driving behaviors and route conditions affect fuel efficiency. The results indicate that both individual driving habits and route characteristics have a significant influence on fuel efficiency.

Zhipeng Ma, Bo Nørregaard Jørgensen, Zheng Grace Ma

Advanced machine learning algorithms are increasingly utilized to provide data-based prediction and decision-making support in Industry 4.0. However, the prediction accuracy achieved by the existing models is insufficient to warrant practical implementation in real-world applications. This is because not all features present in real-world datasets possess a direct relevance to the predictive analysis being conducted. Consequently, the careful incorporation of select features has the potential to yield a substantial positive impact on the outcome. To address the research gap, this paper proposes a novel hybrid framework that combines the feature importance detector - local interpretable model-agnostic explanations (LIME) and the feature interaction detector - neural interaction detection (NID), to improve prediction accuracy. By applying the proposed framework, unnecessary features can be eliminated, and interactions are encoded to generate a more conducive dataset for predictive purposes. Subsequently, the proposed model is deployed to refine the prediction of electricity consumption in foundry processing. The experimental outcomes reveal an augmentation of up to 9.56% in the R2 score, and a diminution of up to 24.05% in the root mean square error.

Zhipeng Ma, Ali Rida Bahja, Andreas Burgdorf et al.

Enhancing fuel efficiency in public transportation requires the integration of complex multimodal data into interpretable, decision-relevant insights. However, traditional analytics and visualization methods often yield fragmented outputs that demand extensive human interpretation, limiting scalability and consistency. This study presents a multi-agent framework that leverages multimodal large language models (LLMs) to automate data narration and energy insight generation. The framework coordinates three specialized agents, including a data narration agent, an LLM-as-a-judge agent, and an optional human-in-the-loop evaluator, to iteratively transform analytical artifacts into coherent, stakeholder-oriented reports. The system is validated through a real-world case study on public bus transportation in Northern Jutland, Denmark, where fuel efficiency data from 4006 trips are analyzed using Gaussian Mixture Model clustering. Comparative experiments across five state-of-the-art LLMs and three prompting paradigms identify GPT-4.1 mini with Chain-of-Thought prompting as the optimal configuration, achieving 97.3% narrative accuracy while balancing interpretability and computational cost. The findings demonstrate that multi-agent orchestration significantly enhances factual precision, coherence, and scalability in LLM-based reporting. The proposed framework establishes a replicable and domain-adaptive methodology for AI-driven narrative generation and decision support in energy informatics.

Zhipeng Ma, Bo Nørregaard Jørgensen, Zheng Grace Ma

Industrial process monitoring increasingly relies on sensor-generated time-series data, yet the lack of labels, high variability, and operational noise make it difficult to extract meaningful patterns using conventional methods. Existing clustering techniques either rely on fixed distance metrics or deep models designed for static data, limiting their ability to handle dynamic, unstructured industrial sequences. Addressing this gap, this paper proposes a novel framework for unsupervised discovery of operational modes in univariate time-series data using image-based convolutional clustering with composite internal evaluation. The proposed framework improves upon existing approaches in three ways: (1) raw time-series sequences are transformed into grayscale matrix representations via overlapping sliding windows, allowing effective feature extraction using a deep convolutional autoencoder; (2) the framework integrates both soft and hard clustering outputs and refines the selection through a two-stage strategy; and (3) clustering performance is objectively evaluated by a newly developed composite score, S_eva, which combines normalized Silhouette, Calinski-Harabasz, and Davies-Bouldin indices. Applied to over 3900 furnace melting operations from a Nordic foundry, the method identifies seven explainable operational patterns, revealing significant differences in energy consumption, thermal dynamics, and production duration. Compared to classical and deep clustering baselines, the proposed approach achieves superior overall performance, greater robustness, and domain-aligned explainability. The framework addresses key challenges in unsupervised time-series analysis, such as sequence irregularity, overlapping modes, and metric inconsistency, and provides a generalizable solution for data-driven diagnostics and energy optimization in industrial systems.

Zhipeng Ma, Bo Nørregaard Jørgensen, Zheng Grace Ma

Improving energy efficiency in industrial foundry processes is a critical challenge, as these operations are highly energy-intensive and marked by complex interdependencies among process variables. Correlation-based analyses often fail to distinguish true causal drivers from spurious associations, limiting their usefulness for decision-making. This paper applies a time-series causal inference framework to identify the operational factors that directly affect energy efficiency in induction furnace melting. Using production data from a Danish foundry, the study integrates time-series clustering to segment melting cycles into distinct operational modes with the PCMCI+ algorithm, a state-of-the-art causal discovery method, to uncover cause-effect relationships within each mode. Across clusters, robust causal relations among energy consumption, furnace temperature, and material weight define the core drivers of efficiency, while voltage consistently influences cooling water temperature with a delayed response. Cluster-specific differences further distinguish operational regimes: efficient clusters are characterized by stable causal structures, whereas inefficient ones exhibit reinforcing feedback loops and atypical dependencies. The contributions of this study are twofold. First, it introduces an integrated clustering-causal inference pipeline as a methodological innovation for analyzing energy-intensive processes. Second, it provides actionable insights that enable foundry operators to optimize performance, reduce energy consumption, and lower emissions.