Shengxing Zhang

About

Shengxing Zhang is an Associate Professor of Economics at the Tepper School of Business, Carnegie Mellon University. His academic role involves research and teaching within the field of economics, with a focus on areas related to business, management science, and organizational behavior. As part of the Tepper School's faculty, he contributes to the school's strategic vision of integrating business, technology, and analytics, supporting the school's mission to lead in these interdisciplinary areas.

Research topics

• Computer Science
• Artificial Intelligence
• Natural Language Processing
• Information Retrieval
• Data Mining
• Theoretical computer science
• Machine Learning
• Programming language

Selected publications

• Supportiveness-based Knowledge Rewriting for Retrieval-augmented Language Modeling

  2025-01-01 · 2 citations

  articleOpen accessSenior author

  Retrieval-augmented language models (RALMs) have recently shown great potential in mitigating the limitations of implicit knowledge in LLMs, such as untimely updating of the latest expertise and unreliable retention of long-tail knowledge.However, since the external knowledge base, as well as the retriever, can not guarantee reliability, potentially leading to the knowledge retrieved not being helpful or even misleading for LLM generation.In this paper, we introduce Supportiveness-based Knowledg…

  PublisherOA PDFDOI

• RewardAnything: Generalizable Principle-Following Reward Models

  ArXiv.org · 2025-06-04 · 1 citations

  preprintOpen access

  Reward Models, essential for guiding Large Language Model optimization, are typically trained on fixed preference datasets, resulting in rigid alignment to single, implicit preference distributions. This prevents adaptation to diverse real-world needs-from conciseness in one task to detailed explanations in another. The standard practice of collecting task-specific preference data and retraining reward models is resource-intensive, often producing biased rewards, and limits practical application…

  PublisherOA PDFDOI

• SampleMix: A Sample-wise Pre-training Data Mixing Strategey by Coordinating Data Quality and Diversity

  ArXiv.org · 2025-03-03

  preprintOpen access

  Existing pretraining data mixing methods for large language models (LLMs) typically follow a domain-wise methodology, a top-down process that first determines domain weights and then performs uniform data sampling across each domain. However, these approaches neglect significant inter-domain overlaps and commonalities, failing to control the global diversity of the constructed training dataset. Further, uniform sampling within domains ignores fine-grained sample-specific features, potentially le…

  PublisherOA PDFDOI

• 3D Surface Reconstruction with Enhanced High-Frequency Details

  ArXiv.org · 2025-05-06

  preprintOpen access1st authorCorresponding

  Neural implicit 3D reconstruction can reproduce shapes without 3D supervision, and it learns the 3D scene through volume rendering methods and neural implicit representations. Current neural surface reconstruction methods tend to randomly sample the entire image, making it difficult to learn high-frequency details on the surface, and thus the reconstruction results tend to be too smooth. We designed a method (FreNeuS) based on high-frequency information to solve the problem of insufficient surfa…

  PublisherOA PDFDOI

• AutoRed: A Free-form Adversarial Prompt Generation Framework for Automated Red Teaming

  ArXiv.org · 2025-10-09

  preprintOpen access

  The safety of Large Language Models (LLMs) is crucial for the development of trustworthy AI applications. Existing red teaming methods often rely on seed instructions, which limits the semantic diversity of the synthesized adversarial prompts. We propose AutoRed, a free-form adversarial prompt generation framework that removes the need for seed instructions. AutoRed operates in two stages: (1) persona-guided adversarial instruction generation, and (2) a reflection loop to iteratively refine low-…

  PublisherOA PDFDOI

Frequent coauthors

• Wei Ye

  Peking University

  46 shared

• Rui Xie

  Peking University

  21 shared

• Jinan Sun

  Peking University

  20 shared

• Wei Ye

  19 shared

• Lifu Wang

  Mianyang Third People's Hospital

  17 shared

• Wen Zhao

  15 shared

• Wei Ye

  14 shared

• Wei Ye

  14 shared

Similar researchers at Carnegie Mellon University

• Denise Rousseauh-76
• Robert Millerh-67
• David Krackhardth-54