SymPyBench: A Dynamic Benchmark for Scientific Reasoning with Executable Python Code

It is our great pleasure to welcome you to the 25th International World Wide Web Conference, WWW 2016. For more than two decades, WWW has been the premier venue for researchers, academics, businesses, and standard bodies to come together and discuss latest updates on the state and evolutionary path of the Web. The conference program of WWW 2016 consists of 11 tracks for refereed research papers, as well as a poster track, a demonstration track, a developers track, an industry track, a Ph.D. symposium track, panels, tutorials, workshops, and a W3C track. Several co-located and special events will also be part of the general program. As in previous years, updates and changes were made to ensure that track topics correctly reflect the key research areas of interest in the WWW community. The research tracks are: Behavioral Analysis and PersonalizationContent AnalysisCrowdsourcing Systems and Social MediaEconomics and MarketsMobilitySecurity and PrivacySemantics and Big DataSocial Networks and Graph AnalysisWeb MiningWeb ScienceWeb Search Systems and Applications The research tracks attracted submissions from all over the world. Submissions were rigorously reviewed, with each paper receiving at least three reviews. During a discussion phase, reviewers and track chairs discussed reviews in detail. For consistency across tracks, final decisions on papers were made at a face-toface PC meeting of all track chairs and the PC chairs. Of the 727 papers submitted to the research tracks, 115 were accepted for an acceptance rate of less than 16%.

Subhadip Nandi, Tanishka Agarwal, Anshika Singh, Priyanka Bhatt. Proceedings of the 19th Conference of the European Chapter of the Association for Computational Linguistics (Volume 5: Industry Track). 2026.

Anagha Kulkarni, Parin Rajesh Jhaveri, Prasha Shrestha, Yu Tong Han, Reza Amini, Behrouz Madahian. Proceedings of the 19th Conference of the European Chapter of the Association for Computational Linguistics (Volume 5: Industry Track). 2026.

Jisoo Mok, Mohammad Kachuee, Shuyang Dai, Shayan Ray, Tara Taghavi, Sungroh Yoon. Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 6: Industry Track). 2024.

Anubhav Shrimal, Avi Jain, Kartik Mehta, Promod Yenigalla. Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies: Industry Track. 2022.

Biplob Biswas, Renhao Cui, Rajiv Ramnath. Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies: Industry Track. 2022.

Joanna Bitton, Maya Pavlova, Ivan Evtimov. Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies: Industry Track. 2022.

Kasturi Bhattacharjee, Rashmi Gangadharaiah, Kathleen McKeown, Dan Roth. Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies: Industry Track. 2022.

As the overlap between traditional computational mechanics and machine learning grows, there is an increasing demand for straight-forward approaches to interface Python-based procedures with C++-based OpenFOAM. This article introduces one such general methodology, allowing the execution of Python code directly within an OpenFOAM solver without the need for Python code translation. The proposed approach is based on the lightweight library pybind11, where OpenFOAM data is transferred to an embedded Python interpreter for manipulation, and results are returned as needed. Following a review of related approaches, the article describes the approach, with a particular focus on data transfer between Python and OpenFOAM, executing Python scripts and functions, and practical details about the implementation in OpenFOAM. Three complementary test cases are presented to highlight the functionality and demonstrate the effect of different data transfer approaches: a Python-based velocity profile boundary condition; a Python-based solver for prototyping; and a machine learning mechanical constitutive law class for solids4foam which performs field calculations.

Parallel python with dask and jupyter The dask framework provides an incredibly useful environment for parallel execution of python code in interactive settings (e.g. jupyter) or batch mode. Its key features are (from what I’ve seen so far): Representation of threading, multiprocessing, and distributed computing with one unified API and CLI.

COLING is held every other year under the auspices of ICCL.This conference is to discuss theoretical and practical problems occurring when languages are handled by computer. The discussion covers a wide range of linguistic issues and their computer implementation. These include: (a) linguistic theories such as syntax, semantics, and discourse problems; (b) linguistic data such as dictionaries and text corpora; (c) algorithms for analysis and synthesis; (d) experimental systems for language understanding and dialogue; and (e) application systems such as those involving machine translation and man-machine interface. This conference provides the bases and technologies essential for the future information society.The 1st International Conference was held in New York in 1965, and the 8th Conference was in Tokyo in 1980. The 15th Conference in Kyoto will be a good opportunity to learn what role languages will play in the information- and multimedia-oriented society. We expect this Conference to promote greater development in this field.

We are delighted to join Ari Juels in welcoming you to the 13th ACM Conference on Computer and Communications Security, held October 30 to November 3, 2006 at the Hilton Alexandria Mark Center, Alexandria, Virginia, USA. These proceedings contain 38 papers presented in the research track and a paper contributed by our keynote speaker, Peter Neumann. The conference also comprises an industry track, tutorials, and workshops.The 38 research track papers contained in this proceedings were selected from 256 received submissions. These submissions were carefully reviewed by at least three members of the program committee, or four in the case of papers authored by program committee members. Reviewing was double-blind, meaning that the program committee was not able to see the names and affiliations of the authors, and the authors were not told which committee members reviewed which papers. Program committee members were allowed to use external reviewers, but were responsible for the contents of the review and representing papers during the decision making. The review phase was followed by a three-week discussion phase in which each paper with at least one supporting review was discussed, outside experts were consulted where needed, and final decisions were made.

On behalf of the entire Organizing Committee it is our great pleasure to welcome you to the 8th joint meeting of the European Software Engineering Conference and the ACM SIGSOFT Symposium on the Foundations of Software Engineering, an internationally renowned forum for researchers, practitioners, and educators to present and discuss the most recent innovations, trends, experiences, and challenges in the field of software engineering. Every second year FSE is jointly organized with ESEC on the Old Continent of Europe. This year's host city is Szeged in the South of Hungary. Szeged, which has always been referred to as the city of sunshine, is simply beautiful and has an atmosphere that no other city has in Hungary. Its sights, like the Votive church, the Synagogue, the Hero's Arch, and the University of Szeged all give a distinct character to the town. From September until late June, the local cafés, restaurants, shops, and streets are all noisy from the lively bustle of university students and the whole town is turned into a nest of youthful spirit. Let us invite you to be part of this unique atmosphere and enjoy your stay during the ESEC/FSE 2011 Conference. This year's program promises to be very exciting. The research track call for papers attracted 203 papers from all continents. After extensive virtual and physical discussions, the Program Committee accepted 34 papers covering a large veriety of software engineering topics -- ranging from mining software archives and empirical studies over program analysis and testing to collaboration, models, and requirements. Three papers received an ACM SIGSOFT Distinguished Paper Award; they are marked in the table of contents. With this ESEC/FSE, we especially wanted to encourage authors to provide artifacts -- tools or data sets that would help others to build on earlier work. A third of the authors of accepted papers submitted artifacts, judged and assessed by an Artifact Evaluation Board. Authors were enthusiastic about participating and eager to improve their packages, further motivated by a 1,000 US$ Best Artifact Award from Microsoft Research. Overall, seven artifacts were found to meet or exceed expectations, and thus are especially recommended by the Board as a base for future research; you will find them marked in the table of contents. The whole-day industrial track is another newly added element of the ESEC/FSE program with the purpose to facilitate an open exchange of ideas between academic researchers and practitioners in industry. ESEC/FSE 2011 will also feature Technical Briefings, an all-day event for communicating the state of topics related to software engineering. This year's conference also offers a comprehensive student program, including the Doctoral Symposium, the Technical Briefings and the PhD Working Groups (PWGs). PWGs are offered for PhD students and other young (pre-doc) researchers with the aim of providing them with the opportunity to meet prominent senior researchers in certain software engineering fields and work with them throughout the conference week.

Mathematical problem solving remains a significant challenge for large language models (LLMs) due to the inherent complexity of mathematical reasoning and the precision required in calculations. This paper introduces a novel approach that leverages Tool-Integrated Reasoning (TIR) and Python code execution to address these challenges and enhance the mathematical reasoning capabilities of LLMs. We fine-tune the DeepSeekMath-Base 7B model using a two-stage process. The first stage involves training on a diverse dataset of natural language mathematical problems using a Chain of Thought (CoT) template, improving the model’s understanding of reasoning paths. In the second stage, we introduce a synthetic dataset that includes examples of tool-integrated reasoning, enabling the model to generate and execute Python code as part of its problem-solving process. Our results demonstrate significant performance improvements, with the model achieving an accuracy of 0.7782 and a maj@N score of 0.7344, outperforming existing models in the domain. This research not only advances the state of mathematical reasoning in LLMs but also provides a framework for integrating external tools to improve reasoning and decision-making processes in AI systems.

This paper presents a novel approach to enhancing educational question-answering (Q&A) systems by combining Retrieval-Augmented Generation (RAG) with Large Language Model (LLM) Code Interpreters. Traditional educational Q&A systems face challenges in areas such as knowledge updates, reasoning accuracy, and the handling of complex computational tasks. These limitations are particularly evident in domains requiring multi-step reasoning or access to real-time, domain-specific knowledge. To address these issues, we propose a system that utilizes RAG to dynamically retrieve up-to-date, relevant information from external knowledge sources, thus mitigating the common "hallucination" problem in LLMs. Additionally, the integration of an LLM Code Interpreter enables the system to perform multi-step logical reasoning and execute Python code for precise calculations, significantly improving its ability to solve mathematical problems and handle complex queries. We evaluated our proposed system on five educational datasets-AI2_ARC, OpenBookQA, E-EVAL, TQA, and ScienceQA-which represent diverse question types and domains. Compared to vanilla Large Language Models (LLMs), our approach combining Retrieval-Augmented Generation (RAG) with Code Interpreters achieved an average accuracy improvement of 10-15 percentage points. Among tested models, GPT-4o and Gemini-pro-1.5 consistently showed the strongest performance, excelling particularly in scientific reasoning, multi-step computations. Despite these advancements, we identify several challenges that remain, including knowledge retrieval failures, code execution errors, difficulties in synthesizing cross-disciplinary information, and limitations in multi-modal reasoning, particularly when combining text and images. These challenges provide important directions for future research aimed at further optimizing educational Q&A systems. Our work shows that integrating RAG and Code Interpreters offers a promising path toward more accurate, transparent, and personalized educational Q&A systems, and can significantly improve the learning experience in various educational contexts.