大模型对齐与安全 - arXivDaily 专题

2501.03991 2025-01-08 cs.CL 57%

Influences on LLM Calibration: A Study of Response Agreement, Loss Functions, and Prompt Styles

Yuxi Xia, Pedro Henrique Luz de Araujo, Klim Zaporojets, Benjamin Roth

专题命中幻觉与事实性：alignment(abstract)；分类 cs.CL

Comments 24 pages, 11 figures, 8 tables

2501.02699 2025-01-07 cs.CV cs.AI 57%

EAGLE: Enhanced Visual Grounding Minimizes Hallucinations in Instructional Multimodal Models

Andrés Villa, Juan León Alcázar, Motasem Alfarra, Vladimir Araujo, Alvaro Soto, Bernard Ghanem

专题命中幻觉与事实性：alignment(abstract)；分类 cs.AI

Comments 12 pages, 4 figures, 8 tables

2412.18537 2025-01-07 cs.CL 57%

Harnessing Large Language Models for Knowledge Graph Question Answering via Adaptive Multi-Aspect Retrieval-Augmentation

Derong Xu, Xinhang Li, Ziheng Zhang, Zhenxi Lin, Zhihong Zhu, Zhi Zheng, Xian Wu, Xiangyu Zhao, Tong Xu, Enhong Chen

专题命中幻觉与事实性：alignment(abstract)；分类 cs.CL

Comments Accepted by AAAI'2025

2501.00907 2025-01-03 cs.SD cs.CL eess.AS 57%

U-GIFT: Uncertainty-Guided Firewall for Toxic Speech in Few-Shot Scenario

Jiaxin Song, Xinyu Wang, Yihao Wang, Yifan Tang, Ru Zhang, Jianyi Liu, Gongshen Liu

专题命中幻觉与事实性：safety(abstract)；分类 cs.CL

Comments 16 pages, 6 figures and 10 tables. Comments are welcome

2412.18404 2024-12-25 cs.CV cs.LG 57%

Extract Free Dense Misalignment from CLIP

JeongYeon Nam, Jinbae Im, Wonjae Kim, Taeho Kil

专题命中幻觉与事实性：alignment(abstract)；分类 cs.LG

Comments 16 pages, 14 figures, AAAI 2025

2412.18051 2024-12-25 cs.CL 57%

Factuality or Fiction? Benchmarking Modern LLMs on Ambiguous QA with Citations

Maya Patel, Aditi Anand

专题命中幻觉与事实性：trustworthy(abstract)；分类 cs.CL

2412.15948 2024-12-23 cs.SE cs.AI cs.HC 57%

Trust Calibration in IDEs: Paving the Way for Widespread Adoption of AI Refactoring

Markus Borg

专题命中幻觉与事实性：trustworthy(abstract)；分类 cs.AI

Comments Accepted for publication in the Proc. of the 2nd Workshop on Integrated Development Environments, 2025

2412.15271 2024-12-23 cs.CL cs.IR 57%

A MapReduce Approach to Effectively Utilize Long Context Information in Retrieval Augmented Language Models

Gongbo Zhang, Zihan Xu, Qiao Jin, Fangyi Chen, Yilu Fang, Yi Liu, Justin F. Rousseau, Ziyang Xu, Zhiyong Lu, Chunhua Weng, Yifan Peng

专题命中幻觉与事实性：safety(abstract)；分类 cs.CL

2409.19454 2024-12-16 cs.HC cs.AI cs.CV 57%

See Where You Read with Eye Gaze Tracking and Large Language Model

Sikai Yang, Gang Yan, Wan Du

专题命中幻觉与事实性：alignment(abstract)；分类 cs.AI

Comments 9 pages

2412.04424 2024-12-06 cs.CV cs.AI 57%

Florence-VL: Enhancing Vision-Language Models with Generative Vision Encoder and Depth-Breadth Fusion

Jiuhai Chen, Jianwei Yang, Haiping Wu, Dianqi Li, Jianfeng Gao, Tianyi Zhou, Bin Xiao

专题命中幻觉与事实性：alignment(abstract)；分类 cs.AI

2403.15263 2024-12-06 cs.LG stat.ML 57%

Federated Bayesian Deep Learning: The Application of Statistical Aggregation Methods to Bayesian Models

John Fischer, Marko Orescanin, Justin Loomis, Patrick McClure

专题命中幻觉与事实性：safety(abstract)；分类 cs.LG

Comments 22 pages, 9 figures

Journal ref IEEE Access (2024)

2410.23178 2024-12-03 astro-ph.IM cs.LG 57%

Uncertainty quantification for fast reconstruction methods using augmented equivariant bootstrap: Application to radio interferometry

Mostafa Cherif, Tobías I. Liaudat, Jonathan Kern, Christophe Kervazo, Jérôme Bobin

专题命中幻觉与事实性：trustworthy(abstract)；分类 cs.LG

Comments 14 pages, 7 figures. Accepted at the Machine Learning and the Physical Sciences Workshop, NeurIPS 2024

2411.09393 2024-11-15 cs.LG 57%

Inherently Interpretable and Uncertainty-Aware Models for Online Learning in Cyber-Security Problems

Benjamin Kolicic, Alberto Caron, Chris Hicks, Vasilios Mavroudis

专题命中幻觉与事实性：trustworthy(abstract)；分类 cs.LG

2411.03497 2024-11-07 cs.CL 57%

Uncertainty Quantification for Clinical Outcome Predictions with (Large) Language Models

Zizhang Chen, Peizhao Li, Xiaomeng Dong, Pengyu Hong

专题命中幻觉与事实性：safety(abstract)；分类 cs.CL

2312.08248 2024-11-06 cs.AI 57%

A Survey of Generative AI for Intelligent Transportation Systems: Road Transportation Perspective

Huan Yan, Yong Li

专题命中幻觉与事实性：safety(abstract)；分类 cs.AI

Comments Revised version submitted to ACM CSUR

2410.14155 2024-11-04 cs.CL 57%

Towards Faithful Natural Language Explanations: A Study Using Activation Patching in Large Language Models

Wei Jie Yeo, Ranjan Satapathy, Erik Cambria

专题命中幻觉与事实性：alignment(abstract)；分类 cs.CL

Comments Under review

2406.01563 2024-11-01 cs.CL 57%

LoFiT: Localized Fine-tuning on LLM Representations

Fangcong Yin, Xi Ye, Greg Durrett

专题命中幻觉与事实性：alignment(abstract)；分类 cs.CL

Comments NeurIPS 2024 Camera Ready

2410.04241 2024-10-30 cs.CL 57%

Adaptive Question Answering: Enhancing Language Model Proficiency for Addressing Knowledge Conflicts with Source Citations

Sagi Shaier, Ari Kobren, Philip Ogren

专题命中幻觉与事实性：trustworthy(abstract)；分类 cs.CL

Comments Accepted to EMNLP 2024

2410.20199 2024-10-29 cs.AI 57%

Rethinking the Uncertainty: A Critical Review and Analysis in the Era of Large Language Models

Mohammad Beigi, Sijia Wang, Ying Shen, Zihao Lin, Adithya Kulkarni, Jianfeng He, Feng Chen, Ming Jin, Jin-Hee Cho, Dawei Zhou, Chang-Tien Lu, Lifu Huang

专题命中幻觉与事实性：safety(abstract)；分类 cs.AI

2402.04957 2024-10-24 cs.CL 57%

Reconfidencing LLMs from the Grouping Loss Perspective

Lihu Chen, Alexandre Perez-Lebel, Fabian M. Suchanek, Gaël Varoquaux

专题命中幻觉与事实性：alignment(abstract)；分类 cs.CL

Comments EMNLP 2024 Findings

2410.16397 2024-10-23 cs.AI cs.ET 57%

Towards a Reliable Offline Personal AI Assistant for Long Duration Spaceflight

Oliver Bensch, Leonie Bensch, Tommy Nilsson, Florian Saling, Wafa M. Sadri, Carsten Hartmann, Tobias Hecking, J. Nathan Kutz

专题命中幻觉与事实性：safety(abstract)；分类 cs.AI

Comments 75th International Astronautical Congress (IAC), Milan, Italy, 14-18 October 2024

2410.15926 2024-10-22 cs.CV cs.CL 57%

Mitigating Object Hallucination via Concentric Causal Attention

Yun Xing, Yiheng Li, Ivan Laptev, Shijian Lu

专题命中幻觉与事实性：alignment(abstract)；分类 cs.CL

Comments To appear at NeurIPS 2024. Code is available at https://github.com/xing0047/cca-llava

2403.19381 2024-10-22 stat.ML cs.LG 57%

On Uncertainty Quantification for Near-Bayes Optimal Algorithms

Ziyu Wang, Chris Holmes

专题命中幻觉与事实性：safety(abstract)；分类 cs.LG

2410.14010 2024-10-21 cs.LG 57%

Conformal Prediction for Federated Graph Neural Networks with Missing Neighbor Information

Ömer Faruk Akgül, Rajgopal Kannan, Viktor Prasanna

专题命中幻觉与事实性：safety(abstract)；分类 cs.LG

2405.07726 2024-10-17 cs.CL 57%

Quantifying and Optimizing Global Faithfulness in Persona-driven Role-playing

Letian Peng, Jingbo Shang

专题命中幻觉与事实性：DPO(abstract)；分类 cs.CL

Comments NeurIPS2024

2406.16562 2024-10-11 cs.CV cs.CL 57%

EVALALIGN: Supervised Fine-Tuning Multimodal LLMs with Human-Aligned Data for Evaluating Text-to-Image Models

Zhiyu Tan, Xiaomeng Yang, Luozheng Qin, Mengping Yang, Cheng Zhang, Hao Li

专题命中幻觉与事实性：alignment(abstract)；分类 cs.CL

Comments Project page: https://sais-fuxi.github.io/projects/evalalign/

2402.11035 2024-10-07 cs.CL cs.IR 57%

Dense Passage Retrieval: Is it Retrieving?

Benjamin Reichman, Larry Heck

专题命中幻觉与事实性：alignment(abstract)；分类 cs.CL

2410.01064 2024-10-03 cs.AI 57%

Truth or Deceit? A Bayesian Decoding Game Enhances Consistency and Reliability

Weitong Zhang, Chengqi Zang, Bernhard Kainz

专题命中幻觉与事实性：alignment(abstract)；分类 cs.AI

2405.00871 2024-09-30 eess.SY cs.LG cs.SY 57%

Learning to Boost the Performance of Stable Nonlinear Systems

Luca Furieri, Clara Lucía Galimberti, Giancarlo Ferrari-Trecate

专题命中幻觉与事实性：safety(abstract)；分类 cs.LG

Journal ref IEEE Open Journal of Control Systems (Volume: 3), 2024

2405.16908 2024-09-27 cs.CL 57%

Can Large Language Models Faithfully Express Their Intrinsic Uncertainty in Words?

Gal Yona, Roee Aharoni, Mor Geva

专题命中幻觉与事实性：alignment(abstract)；分类 cs.CL

Comments To appear in EMNLP 2024 (main conference)

1. 幻觉与事实性 1738 篇

Influences on LLM Calibration: A Study of Response Agreement, Loss Functions, and Prompt Styles

EAGLE: Enhanced Visual Grounding Minimizes Hallucinations in Instructional Multimodal Models

Harnessing Large Language Models for Knowledge Graph Question Answering via Adaptive Multi-Aspect Retrieval-Augmentation

U-GIFT: Uncertainty-Guided Firewall for Toxic Speech in Few-Shot Scenario

Extract Free Dense Misalignment from CLIP

Factuality or Fiction? Benchmarking Modern LLMs on Ambiguous QA with Citations

Trust Calibration in IDEs: Paving the Way for Widespread Adoption of AI Refactoring

A MapReduce Approach to Effectively Utilize Long Context Information in Retrieval Augmented Language Models

See Where You Read with Eye Gaze Tracking and Large Language Model

Florence-VL: Enhancing Vision-Language Models with Generative Vision Encoder and Depth-Breadth Fusion

Federated Bayesian Deep Learning: The Application of Statistical Aggregation Methods to Bayesian Models

Uncertainty quantification for fast reconstruction methods using augmented equivariant bootstrap: Application to radio interferometry

Inherently Interpretable and Uncertainty-Aware Models for Online Learning in Cyber-Security Problems

Uncertainty Quantification for Clinical Outcome Predictions with (Large) Language Models

A Survey of Generative AI for Intelligent Transportation Systems: Road Transportation Perspective

Towards Faithful Natural Language Explanations: A Study Using Activation Patching in Large Language Models

LoFiT: Localized Fine-tuning on LLM Representations

Adaptive Question Answering: Enhancing Language Model Proficiency for Addressing Knowledge Conflicts with Source Citations

Rethinking the Uncertainty: A Critical Review and Analysis in the Era of Large Language Models

Reconfidencing LLMs from the Grouping Loss Perspective

Towards a Reliable Offline Personal AI Assistant for Long Duration Spaceflight

Mitigating Object Hallucination via Concentric Causal Attention

On Uncertainty Quantification for Near-Bayes Optimal Algorithms

Conformal Prediction for Federated Graph Neural Networks with Missing Neighbor Information

Quantifying and Optimizing Global Faithfulness in Persona-driven Role-playing

EVALALIGN: Supervised Fine-Tuning Multimodal LLMs with Human-Aligned Data for Evaluating Text-to-Image Models

Dense Passage Retrieval: Is it Retrieving?

Truth or Deceit? A Bayesian Decoding Game Enhances Consistency and Reliability

Learning to Boost the Performance of Stable Nonlinear Systems

Can Large Language Models Faithfully Express Their Intrinsic Uncertainty in Words?