大模型对齐与安全 - arXivDaily 专题

2012.00419 2021-06-18 cs.LG cs.CY 62%

Machine Learning Systems in the IoT: Trustworthiness Trade-offs for Edge Intelligence

Wiebke Toussaint, Aaron Yi Ding

专题命中安全评测：trustworthy(abstract)；分类 cs.CY、cs.LG

Comments In Proceedings of the Second International Conference on Cognitive Machine Intelligence (CogMI 2020)

2106.03379 2021-06-08 cs.CL cs.AI 62%

LAWDR: Language-Agnostic Weighted Document Representations from Pre-trained Models

Hongyu Gong, Vishrav Chaudhary, Yuqing Tang, Francisco Guzmán

专题命中安全评测：alignment(abstract)；分类 cs.CL、cs.AI

2103.11804 2021-06-08 cs.LG cs.CL cs.IR 62%

Detection of fake news on CoViD-19 on Web Search Engines

V. Mazzeo, A. Rapisarda, G. Giuffrida

专题命中安全评测：trustworthy(abstract)；分类 cs.CL、cs.LG

2102.05884 2021-06-08 cs.LG cs.AI stat.ML 62%

OpinionRank: Extracting Ground Truth Labels from Unreliable Expert Opinions with Graph-Based Spectral Ranking

Glenn Dawson, Robi Polikar

专题命中安全评测：trustworthy(abstract)；分类 cs.AI、cs.LG

Comments 8 pages, 5 figures, accepted at IJCNN 2021

2103.12983 2021-06-03 cs.AI cs.LG 62%

Counterfactual Explanation with Multi-Agent Reinforcement Learning for Drug Target Prediction

Tri Minh Nguyen, Thomas P Quinn, Thin Nguyen, Truyen Tran

专题命中安全评测：trustworthy(abstract)；分类 cs.AI、cs.LG

2105.10026 2021-05-24 cs.CL cs.AI cs.CV 62%

Improving Generation and Evaluation of Visual Stories via Semantic Consistency

Adyasha Maharana, Darryl Hannan, Mohit Bansal

专题命中安全评测：alignment(abstract)；分类 cs.CL、cs.AI

Comments NAACL 2021 (16 pages)

2105.06370 2021-05-14 cs.LG cs.CY 62%

Providing Assurance and Scrutability on Shared Data and Machine Learning Models with Verifiable Credentials

Iain Barclay, Alun Preece, Ian Taylor, Swapna K. Radha, Jarek Nabrzyski

专题命中安全评测：trustworthy(abstract)；分类 cs.CY、cs.LG

Comments This is the submitted, pre-peer reviewed version of this paper

2101.02494 2021-03-15 cs.LG cs.AI cs.SE 62%

Corner case data description and detection

Tinghui Ouyang, Vicent Sant Marco, Yoshinao Isobe, Hideki Asoh, Yutaka Oiwa, Yoshiki Seo

专题命中安全评测：safety(abstract)；分类 cs.AI、cs.LG

2102.00714 2021-02-09 cs.LG cs.AI 62%

NeoRL: A Near Real-World Benchmark for Offline Reinforcement Learning

Rongjun Qin, Songyi Gao, Xingyuan Zhang, Zhen Xu, Shengkai Huang, Zewen Li, Weinan Zhang, Yang Yu

专题命中安全评测：safety(abstract)；分类 cs.AI、cs.LG

2101.07685 2021-01-29 cs.LG cs.AI 62%

GLocalX -- From Local to Global Explanations of Black Box AI Models

Mattia Setzu, Riccardo Guidotti, Anna Monreale, Franco Turini, Dino Pedreschi, Fosca Giannotti

专题命中安全评测：trustworthy(abstract)；分类 cs.AI、cs.LG

Comments 27 pages, 2 figures, submitted to "Special Issue on: Explainable AI (XAI) for Web-based Information Processing"

Journal ref Journal of Artificial Intelligence, Volume 294, May 2021, 103457

2010.07487 2021-01-21 cs.AI cs.CY 62%

Formalizing Trust in Artificial Intelligence: Prerequisites, Causes and Goals of Human Trust in AI

Alon Jacovi, Ana Marasović, Tim Miller, Yoav Goldberg

专题命中安全评测：trustworthy(abstract)；分类 cs.AI、cs.CY

Comments Accepted to ACM FAccT 2021

2012.11769 2020-12-23 cs.LG cs.AI 62%

Self-Progressing Robust Training

Minhao Cheng, Pin-Yu Chen, Sijia Liu, Shiyu Chang, Cho-Jui Hsieh, Payel Das

专题命中安全评测：trustworthy(abstract)；分类 cs.AI、cs.LG

Comments Accepted in AAAI2021

2011.09903 2020-11-20 cs.LG cs.AI 62%

Impact of Accuracy on Model Interpretations

Brian Liu, Madeleine Udell

专题命中安全评测：trustworthy(abstract)；分类 cs.AI、cs.LG

Comments 7 pages, 5 figures

2011.04424 2020-11-11 cs.LG cs.AI cs.RO physics.optics 62%

Playing optical tweezers with deep reinforcement learning: in virtual, physical and augmented environments

Matthew Praeger, Yunhui Xie, James A. Grant-Jacob, Robert W. Eason, Ben Mills

专题命中安全评测：safety(abstract)；分类 cs.AI、cs.LG

1909.09962 2020-11-03 cs.CL cs.LG 62%

Adapting Language Models for Non-Parallel Author-Stylized Rewriting

Bakhtiyar Syed, Gaurav Verma, Balaji Vasan Srinivasan, Anandhavelu Natarajan, Vasudeva Varma

专题命中安全评测：alignment(abstract)；分类 cs.CL、cs.LG

Comments Accepted for publication in Main Technical Track at AAAI 20

2006.09526 2020-10-28 cs.CL cs.LG 62%

Cross-lingual Retrieval for Iterative Self-Supervised Training

Chau Tran, Yuqing Tang, Xian Li, Jiatao Gu

专题命中安全评测：alignment(abstract)；分类 cs.CL、cs.LG

Journal ref NeurIPS 2020

2004.14025 2020-10-08 cs.AI cs.CL 62%

Multi-View Attention Network for Visual Dialog

Sungjin Park, Taesun Whang, Yeochan Yoon, Heuiseok Lim

专题命中安全评测：alignment(abstract)；分类 cs.CL、cs.AI

2003.09761 2020-07-15 cs.CY cs.LG physics.soc-ph stat.ML 62%

Smarter Parking: Using AI to Identify Parking Inefficiencies in Vancouver

Devon Graham, Satish Kumar Sarraf, Taylor Lundy, Ali MohammadMehr, Sara Uppal, Tae Yoon Lee, Hedayat Zarkoob, Scott Duke Kominers, Kevin Leyton-Brown

专题命中安全评测：safety(abstract)；分类 cs.CY、cs.LG

Comments All the authors contributed equally. This paper is an outcome of https://www.cs.ubc.ca/~kevinlb/teaching/cs532l%20-%202018-19/index.html. To be submitted to a journal in transportation or urban planning

1802.05027 2020-07-15 cs.LG cs.AI stat.ML 62%

Not to Cry Wolf: Distantly Supervised Multitask Learning in Critical Care

Patrick Schwab, Emanuela Keller, Carl Muroi, David J. Mack, Christian Strässle, Walter Karlen

专题命中安全评测：trustworthy(abstract)；分类 cs.AI、cs.LG

Journal ref Proceedings of the 35th International Conference on Machine Learning, PMLR 80:4518-4527, 2018

2007.00753 2020-07-07 cs.LG cs.AI stat.ML 62%

Opportunities and Challenges in Deep Learning Adversarial Robustness: A Survey

Samuel Henrique Silva, Peyman Najafirad

专题命中安全评测：safety(abstract)；分类 cs.AI、cs.LG

Comments 20 pages, 9 figures, submited to IEEE Transactions on Knowledge and Data Engineering

2006.11371 2020-06-24 cs.CV cs.AI cs.LG 62%

Opportunities and Challenges in Explainable Artificial Intelligence (XAI): A Survey

Arun Das, Paul Rad

专题命中安全评测：trustworthy(abstract)；分类 cs.AI、cs.LG

Comments 24 pages, 20 figures, survey paper, submitting to IEEE

2006.10609 2020-06-19 cs.LG cs.AI stat.ML 62%

The Clever Hans Effect in Anomaly Detection

Jacob Kauffmann, Lukas Ruff, Grégoire Montavon, Klaus-Robert Müller

专题命中安全评测：trustworthy(abstract)；分类 cs.AI、cs.LG

Comments 17 pages, preprint

1910.12908 2020-03-10 cs.RO cs.AI cs.LG 62%

Certified Adversarial Robustness for Deep Reinforcement Learning

Björn Lütjens, Michael Everett, Jonathan P. How

专题命中安全评测：safety(abstract)；分类 cs.AI、cs.LG

Comments Published at Conference on Robot Learning (CoRL) 2019; (v2) contains minor updates to related works; (v3) acknowledged AWS

Journal ref Proceedings of Machine Learning Research (PMLR) Vol. 100, 2019

1912.05100 2019-12-12 cs.LG cs.AI stat.ML 62%

Explainability Fact Sheets: A Framework for Systematic Assessment of Explainable Approaches

Kacper Sokol, Peter Flach

专题命中安全评测：safety(abstract)；分类 cs.AI、cs.LG

Comments Conference on Fairness, Accountability, and Transparency (FAT* '20), January 27-30, 2020, Barcelona, Spain

1911.11253 2019-11-27 cs.LG cs.AI stat.ML 62%

Playing it Safe: Adversarial Robustness with an Abstain Option

Cassidy Laidlaw, Soheil Feizi

专题命中安全评测：safety(abstract)；分类 cs.AI、cs.LG

1911.10735 2019-11-26 cs.LG cs.AI cs.NE cs.PF 62%

CAMUS: A Framework to Build Formal Specifications for Deep Perception Systems Using Simulators

Julien Girard-Satabin, Guillaume Charpiat, Zakaria Chihani, Marc Schoenauer

专题命中安全评测：safety(abstract)；分类 cs.AI、cs.LG

1911.01275 2019-11-05 cs.CY cs.LG cs.SI 62%

Using Arabic Tweets to Understand Drug Selling Behaviors

Wesam Alruwaili, Bradley Protano, Tejasvi Sirigiriraju, Hamed Alhoori

专题命中安全评测：safety(abstract)；分类 cs.CY、cs.LG

1905.09957 2019-10-29 cs.LG cs.AI cs.CR stat.ML 62%

Robust Attribution Regularization

Jiefeng Chen, Xi Wu, Vaibhav Rastogi, Yingyu Liang, Somesh Jha

专题命中安全评测：trustworthy(abstract)；分类 cs.AI、cs.LG

1909.04436 2019-09-11 cs.LG cs.AI stat.AP stat.ML 62%

The Prevalence of Errors in Machine Learning Experiments

Martin Shepperd, Yuchen Guo, Ning Li, Mahir Arzoky, Andrea Capiluppi, Steve Counsell, Giuseppe Destefanis, Stephen Swift, Allan Tucker, Leila Yousefi

专题命中安全评测：trustworthy(abstract)；分类 cs.AI、cs.LG

Comments 20th International Conference on Intelligent Data Engineering and Automated Learning (IDEAL), 14--16 November 2019

1908.02624 2019-08-08 cs.CY cs.AI 62%

A 20-Year Community Roadmap for Artificial Intelligence Research in the US

Yolanda Gil, Bart Selman

专题命中安全评测：trustworthy(abstract)；分类 cs.AI、cs.CY

Comments A Computing Community Consortium (CCC) workshop report, 109 pages

1. 安全评测 9380 篇

Machine Learning Systems in the IoT: Trustworthiness Trade-offs for Edge Intelligence

LAWDR: Language-Agnostic Weighted Document Representations from Pre-trained Models

Detection of fake news on CoViD-19 on Web Search Engines

OpinionRank: Extracting Ground Truth Labels from Unreliable Expert Opinions with Graph-Based Spectral Ranking

Counterfactual Explanation with Multi-Agent Reinforcement Learning for Drug Target Prediction

Improving Generation and Evaluation of Visual Stories via Semantic Consistency

Providing Assurance and Scrutability on Shared Data and Machine Learning Models with Verifiable Credentials

Corner case data description and detection

NeoRL: A Near Real-World Benchmark for Offline Reinforcement Learning

GLocalX -- From Local to Global Explanations of Black Box AI Models

Formalizing Trust in Artificial Intelligence: Prerequisites, Causes and Goals of Human Trust in AI

Self-Progressing Robust Training

Impact of Accuracy on Model Interpretations

Playing optical tweezers with deep reinforcement learning: in virtual, physical and augmented environments

Adapting Language Models for Non-Parallel Author-Stylized Rewriting

Cross-lingual Retrieval for Iterative Self-Supervised Training

Multi-View Attention Network for Visual Dialog

Smarter Parking: Using AI to Identify Parking Inefficiencies in Vancouver

Not to Cry Wolf: Distantly Supervised Multitask Learning in Critical Care

Opportunities and Challenges in Deep Learning Adversarial Robustness: A Survey

Opportunities and Challenges in Explainable Artificial Intelligence (XAI): A Survey

The Clever Hans Effect in Anomaly Detection

Certified Adversarial Robustness for Deep Reinforcement Learning

Explainability Fact Sheets: A Framework for Systematic Assessment of Explainable Approaches

Playing it Safe: Adversarial Robustness with an Abstain Option

CAMUS: A Framework to Build Formal Specifications for Deep Perception Systems Using Simulators

Using Arabic Tweets to Understand Drug Selling Behaviors

Robust Attribution Regularization

The Prevalence of Errors in Machine Learning Experiments

A 20-Year Community Roadmap for Artificial Intelligence Research in the US