High-quality mathematical and logical datasets with verifiable answers are essential for strengthening the reasoning capabilities of large language models (LLMs). While recent data augmentation techniques have facilitated the creation of large-scale benchmarks, existing LLM-generated datasets often suffer from limited reliability, diversity, and scalability. To address these challenges, we introduce PuzzleClone, a formal framework for synthesizing verifiable data at scale using a novel DSL-driven approach. Our approach features three key innovations: (1) encoding seed puzzles into structured logical specifications, (2) generating scalable variants through systematic variable and constraint randomization, and (3) ensuring validity via a reproduction mechanism. Applying PuzzleClone, we construct PC-83K, a benchmark comprising over 83K diverse and programmatically validated puzzles. The generated puzzles span a wide spectrum of difficulty and formats, posing significant challenges to current state-of-the-art models. Experimental results show that post training (SFT and RL) on PC-83K yields substantial improvements not only on the testset but also on various logic and mathematical benchmarks. Post training raises average performance on PC-83K from 14.5 to 66.0 and delivers consistent improvements across 7 logic and mathematical benchmarks up to 18.4 absolute percentage points (SATBench from 51.6 to 70.0). Our code and data are available at https://github.com/HiThink-Research/PuzzleClone.

Large language models (LLMs) are increasingly used as web-facing agents for information-seeking tasks over live websites. Many such tasks require multi-step in-site navigation and cross-page evidence aggregation, a setting known as web traversal. Existing LLM-based web agents, including recent explore–critic architectures, typically store interaction history as an ever-growing unstructured text trace and mix high-level planning with low-level page actions. This makes long-horizon behavior brittle, obscures the source of evidence, and often leads to redundant browsing or premature stopping. To mitigate these issues, we propose WebTraveler, a hierarchical multi-agent framework that separates high-level information-seeking planning from local web exploration and maintains a structured trajectory memory of decision-relevant signals. This design provides compact information interfaces for evidence sufficiency checking and webpage navigation, improving long-horizon stability and cross-page evidence aggregation. To obtain scalable supervision, we further introduce path-guided data synthesis, which uses executable navigation paths to collect trajectories, removes explicit path hints, and filters invalid samples through answer-consistency checks. Experiments show that WebTraveler consistently outperforms previous methods across LLM scales, especially on tasks requiring deeper navigation, with further gains from path-guided supervision. Moreover, WebTraveler also shows generalization to broader web navigation tasks. Code is available at https://github.com/RongchangLi/WebTraveler.

Reasoning capability plays a significantly critical role in the the broad applications of Large Language Models (LLMs). To enhance the reasoning performance of LLMs, diverse Reinforcement Learning (RL)-based fine-tuning approaches have been proposed to address the limited generalization capability of LLMs trained solely via Supervised Fine-Tuning (SFT). Despite their effectiveness, two major limitations hinder the advancement of LLMs. First, vanilla RL-based approaches ignore annotated Chain-of-Thought (CoT) and incorporate unstable reasoning path sampling, which typically results in model collapse, unstable training process, and suboptimal performance. Second, existing SFT approaches generally overemphasize the annotated CoT, potentially leading to performance degradation due to insufficient exploitation of potential CoT. In this paper, we propose a Contrastive learning with annotated CoT-based Reinforced Fine-Tuning approach, i.e., CARFT, to enhance the reasoning performance of LLMs while addressing the aforementioned limitations. Specifically, we propose learning a representation for each CoT. Based on this representation, we design novel contrastive signals to guide the fine-tuning process. Our approach not only fully exploits the available annotated CoT but also stabilizes the fine-tuning procedure by incorporating an additional unsupervised learning signal. We conduct comprehensive experiments and in-depth analysis with three baseline approaches, two foundation models, and two datasets to demonstrate significant advantages of CARFT in terms of robustness, performance (up to 10.15%), and efficiency (up to 30.62%).

As an important application of Artificial Intelligence, legal intelligence has recently attracted the attention of many researchers. Previous works investigated diverse issues like predicting crimes, predicting outcomes of judicial debates, or extracting information/knowledge from various kinds of legal documents. Although many advances have been made, the research on supporting prediction of court judgments remains relatively scarce, while the lack of large-scale data resources limits the development of this research.In this paper, we present a novel, large-size Court Debate Dataset (CDD), which includes 30,481 court cases, totaling 1,144,425 utterances. CDD contains real-world conversations involving judges, plaintiffs and defendants in court trials. To construct this dataset we have invited experienced judges to design appropriate labels for data records. We then asked law school students to provide annotations based on the defined labels. The dataset can be applied to several downstream tasks, such as text summarization, dialogue generation, text classification, etc. We introduce the details of the different tasks in the rapidly developing field of legal intelligence, the research of which can be fostered thanks to our dataset, and we provide the corresponding benchmark performance.

Most recently proposed approaches in dialogue state tracking (DST) leverage the context and the last dialogue states to track current dialogue states, which are often slot-value pairs. Although the context contains the complete dialogue information, the information is usually indirect and even requires reasoning to obtain. The information in the lastly predicted dialogue states is direct, but when there is a prediction error, the dialogue information from this source will be incomplete or erroneous. In this paper, we propose the Dialogue State Tracking with Multi-Level Fusion of Predicted Dialogue States and Conversations network (FPDSC). This model extracts information of each dialogue turn by modeling interactions among each turn utterance, the corresponding last dialogue states, and dialogue slots. Then the representation of each dialogue turn is aggregated by a hierarchical structure to form the passage information, which is utilized in the current turn of DST. Experimental results validate the effectiveness of the fusion network with 55.03% and 59.07% joint accuracy on MultiWOZ 2.0 and MultiWOZ 2.1 datasets, which reaches the state-of-the-art performance. Furthermore, we conduct the deleted-value and related-slot experiments on MultiWOZ 2.1 to evaluate our model.