Jie Fang

Junjie Yu, Pengrui Lu, Weiye Si et al.

Benchmarks within the OpenClaw ecosystem have thus far evaluated exclusively assistant-level tasks, leaving the academic-level capabilities of OpenClaw largely unexamined. We introduce AcademiClaw, a bilingual benchmark of 80 complex, long-horizon tasks sourced directly from university students' real academic workflows -- homework, research projects, competitions, and personal projects -- that they found current AI agents unable to solve effectively. Curated from 230 student-submitted candidates through rigorous expert review, the final task set spans 25+ professional domains, ranging from olympiad-level mathematics and linguistics problems to GPU-intensive reinforcement learning and full-stack system debugging, with 16 tasks requiring CUDA GPU execution. Each task executes in an isolated Docker sandbox and is scored on task completion by multi-dimensional rubrics combining six complementary techniques, with an independent five-category safety audit providing additional behavioral analysis. Experiments on six frontier models show that even the best achieves only a 55\% pass rate. Further analysis uncovers sharp capability boundaries across task domains, divergent behavioral strategies among models, and a disconnect between token consumption and output quality, providing fine-grained diagnostic signals beyond what aggregate metrics reveal. We hope that AcademiClaw and its open-sourced data and code can serve as a useful resource for the OpenClaw community, driving progress toward agents that are more capable and versatile across the full breadth of real-world academic demands. All data and code are available at https://github.com/GAIR-NLP/AcademiClaw.

Jie Fang, Xiongwei Wu, Dianchao Lin et al.

The growing use of probe vehicles generates a huge number of GNSS data. Limited by the satellite positioning technology, further improving the accuracy of map-matching is challenging work, especially for low-frequency trajectories. When matching a trajectory, the ego vehicle's spatial-temporal information of the present trip is the most useful with the least amount of data. In addition, there are a large amount of other data, e.g., other vehicles' state and past prediction results, but it is hard to extract useful information for matching maps and inferring paths. Most map-matching studies only used the ego vehicle's data and ignored other vehicles' data. Based on it, this paper designs a new map-matching method to make full use of "Big data". We first sort all data into four groups according to their spatial and temporal distance from the present matching probe which allows us to sort for their usefulness. Then we design three different methods to extract valuable information (scores) from them: a score for speed and bearing, a score for historical usage, and a score for traffic state using the spectral graph Markov neutral network. Finally, we use a modified top-K shortest-path method to search the candidate paths within an ellipse region and then use the fused score to infer the path (projected location). We test the proposed method against baseline algorithms using a real-world dataset in China. The results show that all scoring methods can enhance map-matching accuracy. Furthermore, our method outperforms the others, especially when GNSS probing frequency is less than 0.01 Hz.

Zilong Bai, Ruiji Zhang, Linqing Chen et al.

In recent years, large language models(LLMs) have attracted significant attention due to their exceptional performance across a multitude of natural language process tasks, and have been widely applied in various fields. However, the application of large language models in the Intellectual Property (IP) domain is challenging due to the strong need for specialized knowledge, privacy protection, processing of extremely long text in this field. In this technical report, we present for the first time a low-cost, standardized procedure for training IP-oriented LLMs, meeting the unique requirements of the IP domain. Using this standard process, we have trained the PatentGPT series models based on open-source pretrained models. By evaluating them on the open-source IP-oriented benchmark MOZIP, our domain-specific LLMs outperforms GPT-4, indicating the effectiveness of the proposed training procedure and the expertise of the PatentGPT models in the IP domain. Remarkably, our model surpassed GPT-4 on the 2019 China Patent Agent Qualification Examination, scoring 65 and matching human expert levels. Additionally, the PatentGPT model, which utilizes the SMoE architecture, achieves performance comparable to that of GPT-4 in the IP domain and demonstrates a better cost-performance ratio on long-text tasks, potentially serving as an alternative to GPT-4 within the IP domain.

Linqing Chen, Weilei Wang, Yubin Xia et al.

In this paper, we propose a novel system that integrates state-of-the-art, domain-specific large language models with advanced information retrieval techniques to deliver comprehensive and context-aware responses. Our approach facilitates seamless interaction among diverse components, enabling cross-validation of outputs to produce accurate, high-quality responses enriched with relevant data, images, tables, and other modalities. We demonstrate the system's capability to enhance response precision by leveraging a robust question-answering model, significantly improving the quality of dialogue generation. The system provides an accessible platform for real-time, high-fidelity interactions, allowing users to benefit from efficient human-computer interaction, precise retrieval, and simultaneous access to a wide range of literature and data. This dramatically improves the research efficiency of professionals in the biomedical and pharmaceutical domains and facilitates faster, more informed decision-making throughout the R\&D process. Furthermore, the system proposed in this paper is available at https://synapse-chat.patsnap.com.

Linqing Chen, Weilei Wang, Zilong Bai et al.

Large language models (LLMs) have revolutionized Natural Language Processing (NLP) by minimizing the need for complex feature engineering. However, the application of LLMs in specialized domains like biopharmaceuticals and chemistry remains largely unexplored. These fields are characterized by intricate terminologies, specialized knowledge, and a high demand for precision areas where general purpose LLMs often fall short. In this study, we introduce PharmaGPT, a suite of domain specilized LLMs with 13 billion and 70 billion parameters, specifically trained on a comprehensive corpus tailored to the Bio-Pharmaceutical and Chemical domains. Our evaluation shows that PharmaGPT surpasses existing general models on specific-domain benchmarks such as NAPLEX, demonstrating its exceptional capability in domain-specific tasks. Remarkably, this performance is achieved with a model that has only a fraction, sometimes just one-tenth-of the parameters of general-purpose large models. This advancement establishes a new benchmark for LLMs in the bio-pharmaceutical and chemical fields, addressing the existing gap in specialized language modeling. It also suggests a promising path for enhanced research and development, paving the way for more precise and effective NLP applications in these areas.

Jie Fang, Jianwu Lin, Shutao Xia et al.

Instead of conducting manual factor construction based on traditional and behavioural finance analysis, academic researchers and quantitative investment managers have leveraged Genetic Programming (GP) as an automatic feature construction tool in recent years, which builds reverse polish mathematical expressions from trading data into new factors. However, with the development of deep learning, more powerful feature extraction tools are available. This paper proposes Neural Network-based Automatic Factor Construction (NNAFC), a tailored neural network framework that can automatically construct diversified financial factors based on financial domain knowledge and a variety of neural network structures. The experiment results show that NNAFC can construct more informative and diversified factors than GP, to effectively enrich the current factor pool. For the current market, both fully connected and recurrent neural network structures are better at extracting information from financial time series than convolution neural network structures. Moreover, new factors constructed by NNAFC can always improve the return, Sharpe ratio, and the max draw-down of a multi-factor quantitative investment strategy due to their introducing more information and diversification to the existing factor pool.

Jie Fang, Jianwu Lin

One of the major characteristics of financial time series is that they contain a large amount of non-stationary noise, which is challenging for deep neural networks. People normally use various features to address this problem. However, the performance of these features depends on the choice of hyper-parameters. In this paper, we propose to use neural networks to represent these indicators and train a large network constructed of smaller networks as feature layers to fine-tune the prior knowledge represented by the indicators. During back propagation, prior knowledge is transferred from human logic to machine logic via gradient descent. Prior knowledge is the deep belief of neural network and teaches the network to not be affected by non-stationary noise. Moreover, co-distillation is applied to distill the structure into a much smaller size to reduce redundant features and the risk of overfitting. In addition, the decisions of the smaller networks in terms of gradient descent are more robust and cautious than those of large networks. In numerical experiments, we find that our algorithm is faster and more accurate than traditional methods on real financial datasets. We also conduct experiments to verify and comprehend the method.

Jie Fang, Shutao Xia, Jianwu Lin et al.

Genetic programming (GP) is the state-of-the-art in financial automated feature construction task. It employs reverse polish expression to represent features and then conducts the evolution process. However, with the development of deep learning, more powerful feature extraction tools are available. This paper proposes Alpha Discovery Neural Network (ADNN), a tailored neural network structure which can automatically construct diversified financial technical indicators based on prior knowledge. We mainly made three contributions. First, we use domain knowledge in quantitative trading to design the sampling rules and object function. Second, pre-training and model pruning has been used to replace genetic programming, because it can conduct more efficient evolution process. Third, the feature extractors in ADNN can be replaced by different feature extractors and produce different functions. The experiment results show that ADNN can construct more informative and diversified features than GP, which can effectively enriches the current factor pool. The fully-connected network and recurrent network are better at extracting information from the financial time series than the convolution neural network. In real practice, features constructed by ADNN can always improve multi-factor strategies' revenue, sharpe ratio, and max draw-down, compared with the investment strategies without these factors.

Jie Fang, Shutao Xia, Jianwu Lin et al.

In automatic financial feature construction task, the state-of-the-art technic leverages reverse polish expression to represent the features, then use genetic programming (GP) to conduct its evolution process. In this paper, we propose a new framework based on neural network, alpha discovery neural network (ADNN). In this work, we made several contributions. Firstly, in this task, we make full use of neural network overwhelming advantage in feature extraction to construct highly informative features. Secondly, we use domain knowledge to design the object function, batch size, and sampling rules. Thirdly, we use pre-training to replace the GP evolution process. According to neural network universal approximation theorem, pre-training can conduct a more effective and explainable evolution process. Experiment shows that ADNN can remarkably produce more diversified and higher informative features than GP. Besides, ADNN can serve as a data augmentation algorithm. It further improves the the performance of financial features constructed by GP.