Evaluating the Promise and Pitfalls of LLMs in Hiring Decisions
Comments 10 pages, 2 figures, 2 tables. Submitted to NeurIPS 2025
期刊&会议
NeurIPS
Conference on Neural Information Processing Systems · 会议 · Machine Learning
Directly Learning Stock Trading Strategies Through Profit Guided Loss Functions
机构 * Rochester Institute of Technology（罗切斯特理工学院）
Comments 17 pages, 4 figures, Submitted to Neural Information Processing Systems 2025
Compositional Abilities Emerge Multiplicatively: Exploring Diffusion Models on a Synthetic Task
机构 * Center for Brain Science, Harvard University（哈佛大学脑科学中心） ； Physics & Informatics Laboratories, NTT Research, Inc.（NTT研究所物理与信息实验室） ； EECS Department, University of Michigan（密歇根大学电子工程与计算机科学系）
Comments 37th Conference on Neural Information Processing Systems (NeurIPS)
In-Trajectory Inverse Reinforcement Learning: Learn Incrementally Before An Ongoing Trajectory Terminates
Journal ref 2024 Advances in Neural Information Processing Systems
ACMP: Allen-Cahn Message Passing with Attractive and Repulsive Forces for Graph Neural Networks
机构 * Shanghai Jiao Tong University（上海交通大学） ； University of New South Wales（新南威尔士大学） ； King Abdullah University of Science and Technology（卡塔尔国王 Abdullah 科学与技术大学） ； Shanghai Jiao Tong University Shanghai Artificial Intelligence Laboratory（上海交通大学上海人工智能实验室）
Comments 26 pages, 5 figures. NeurIPS 2022 Workshop on GLFrontiers (Oral). ICLR 2023 (Spotlight)
From Black Box to Biomarker: Sparse Autoencoders for Interpreting Speech Models of Parkinson's Disease
机构 * McGill University（麦吉尔大学） ； Centre for Research on Brain, Language, and Music（大脑、语言和音乐研究中心） ； Mila - Quebec AI Institute（魁北克人工智能研究所） ； Douglas Research Centre（道格拉斯研究中心） ； Concordia University（Concordia大学）
Comments 14 pages, 5 figures, submitted to NeurIPS 2025
Differentially Private Set Representations
Comments Appears at NeurIPS 2024
LLM world models are mental: Output layer evidence of brittle world model use in LLM mechanical reasoning
Comments Manuscript comprises 14 pages, 4 figures, 4 tables in the Technical Appendix and Supplementary Material, and is under review at NeurIPS 2025
MetaAligner: Towards Generalizable Multi-Objective Alignment of Language Models
机构 * The University of Manchester（曼彻斯特大学）
Comments Accepted by NeurIPS 2024 main track
Journal ref https://proceedings.neurips.cc/paper_files/paper/2024/hash/3d03800841fa1bb2f43ef1750aafcce4-Abstract-Conference.html
Vector Quantization Prompting for Continual Learning
机构 * Communication University of China（中国通信大学） ； Harbin Institute of Technology, Shenzhen（哈尔滨工业大学（深圳）） ； The Chinese University of Hong Kong（香港中文大学）
Comments Accepted by NeurIPS 2024
On Pre-training of Multimodal Language Models Customized for Chart Understanding
Comments NeurIPS 2024 Workshop on Adaptive Foundation Models
ViTally Consistent: Scaling Biological Representation Learning for Cell Microscopy
机构 * Valence Labs（Valence实验室） ； University of Manchester（曼彻斯特大学）
Comments ICML 2025 main-track paper (42nd International Conference on Machine Learning). Formerly appeared as best paper runner-up at NeurIPS 2024 Foundation Models for Science Workshop (38th Conference on Neural Information Processing Systems). 18 pages, 7 figures
Geometry, Computation, and Optimality in Stochastic Optimization
机构 * Stanford University（斯坦福大学）
Comments 47 pages. An earlier version of this paper, entitled "Necessary and Sufficient Geometries for Gradient Methods," appeared at NeurIPS 2019 (arXiv:1909.10455v2)
Large Language Models Engineer Too Many Simple Features For Tabular Data
机构 * University of Freiburg（弗赖堡大学） ； ELLIS Institute Tübingen（图宾根ELLIS研究所）
Comments Accepted at the 3rd Table Representation Learning Workshop @ NeurIPS 2024
Provable Robustness of (Graph) Neural Networks Against Data Poisoning and Backdoor Attacks
机构 * Technical University of Munich（慕尼黑技术大学） ； Munich Data Science Institute（慕尼黑数据科学研究所） ； Munich Center for Machine Learning (MCML)（慕尼黑机器学习中心（MCML）） ； School of Computation, Information and Technology（计算、信息与技术学院）
Comments Published in TMLR. Best Paper Award at the AdvML-Frontiers @ NeurIPS 2024 workshop. Code available at https://github.com/saper0/qpcert
Group-wise oracle-efficient algorithms for online multi-group learning
Comments Updated after NeurIPS 2024 camera-ready
Differentially Private Federated Low Rank Adaptation Beyond Fixed-Matrix
机构 * Department of Electronic Engineering（电子工程系） ； Fudan University（复旦大学） ； School of Computing（计算学院） ； Macquarie University（麦觉瑞大学） ； Department of Communication Science and Engineering（通信科学与工程系）
Comments 23 pages, NeurIPS 2025 under review
Radial Neighborhood Smoothing Recommender System
机构 * Department of Statistics and Bioinformatics and Computational Biology Program（统计学与生物信息学及计算生物学项目部） ； School of Mathematical Sciences and Center for Statistical Science（数学科学学院和统计科学中心）
Comments 34 pages, 2 figures. Submitted to NeurIPS 2025
DESIGN: Encrypted GNN Inference via Server-Side Input Graph Pruning
机构 * University of Notre Dame（诺丁汉大学） ； Florida State University（佛罗里达州立大学） ； University of Central Florida（中央佛罗里达大学）
Comments Under Review in Conference on Neural Information Processing Systems (NeurIPS 2025)
Regularization-based Framework for Quantization-, Fault- and Variability-Aware Training
机构 * Electrical Engineering Department（电子工程系） ； IIT Bombay（印度理工学院班加罗尔）
Comments AB and RS contributed equally to this work. A version of this paper accepted at MLNCP @ NeuRIPS '24
Latent Functional Maps: a spectral framework for representation alignment
机构 * IST Austria（IST奥地利研究院） ； Sapienza, University of Rome（罗马大学萨皮恩扎分校）
Journal ref Advances in Neural Information Processing Systems(NeurIPS) 37 (2024) 66178-66203
Cascade Speculative Drafting for Even Faster LLM Inference
机构 * University of Illinois at Urbana-Champaign（伊利诺伊大学厄巴纳-香槟分校）
Comments NeurIPS 2024
Anchoring AI Capabilities in Market Valuations: The Capability Realization Rate Model and Valuation Misalignment Risk
Comments 11 pages, 3 figures, NeurIPS
Uncovering RL Integration in SSL Loss: Objective-Specific Implications for Data-Efficient RL
Comments RLC 2025, Neurips SSL Workshop 2024
Efficient Multi-Task Reinforcement Learning with Cross-Task Policy Guidance
机构 * Institute of Automation, Chinese Academy of Sciences（中国科学院自动化研究所） ； School of Artificial Intelligence, University of Chinese Academy of Sciences（中国科学院大学人工智能学院） ； Tsinghua University（清华大学） ； Tencent AI Lab（腾讯AI实验室）
Comments NeurIPS2024
Journal ref 38th Conference on Neural Information Processing Systems (NeurIPS 2024)
Advancing Stroke Risk Prediction Using a Multi-modal Foundation Model
机构 * Signal Processing Institute EPFL University Lausanne（瑞士洛桑联邦理工学院信号处理研究所） ； Brigham and Women’s Hospital Harvard Medical School（哈佛医学院布里奇沃特医院） ； Department of Quantitative Biomedicine University of Zurich（苏黎世大学定量生物医学系） ； ETH AI Center, Department of Computer Science Department of Quantitative Biomedicine University of Zurich（苏黎世大学定量生物医学系）
Comments Accepted as oral paper at AIM-FM workshop, Neurips 2024
AbdomenAtlas-8K: Annotating 8,000 CT Volumes for Multi-Organ Segmentation in Three Weeks
机构 * Johns Hopkins University（约翰霍普金斯大学） ； Rutgers University（罗格斯大学） ； City University of Hong Kong（香港城市大学） ； NVIDIA（英伟达）
Comments Conference on Neural Information Processing Systems (NeurIPS 2023)
Computation-Aware Gaussian Processes: Model Selection And Linear-Time Inference
机构 * Columbia University（哥伦比亚大学） ； University of Pennsylvania（宾夕法尼亚大学） ； University of Tübingen, Tübingen AI Center（图宾根大学，图宾根人工智能中心） ； University of British Columbia, Vector Institute（不列颠哥伦比亚大学，向量研究所）
Comments Advances in Neural Information Processing Systems (NeurIPS 2024)
Domain Adaptation of Drag Reduction Policy to Partial Measurements
机构 * Imperial College London（帝国理工学院伦敦分校）
Journal ref Machine Learning and the Physical Sciences Workshop, NeurIPS 2024
Temporal Continual Learning with Prior Compensation for Human Motion Prediction
机构 * Sun Yat-sen University（中山大学） ； Dongguan University of Technology（东莞理工大学）
Comments Advances in Neural Information Processing Systems 2023
Journal ref Advances in Neural Information Processing Systems, 2023, 36: 65837-65849
