大模型对齐与安全

SaDe: Learning Models that Provably Satisfy Domain Constraints

Kshitij Goyal, Sebastijan Dumancic, Hendrik Blockeel

专题命中 其他安全 ：safety(abstract)；分类 cs.AI、cs.LG

Comments 16 pages

ADAPT: Vision-Language Navigation with Modality-Aligned Action Prompts

Bingqian Lin, Yi Zhu, Zicong Chen, Xiwen Liang, Jianzhuang Liu, Xiaodan Liang

专题命中 其他安全 ：alignment(abstract)；分类 cs.CL、cs.AI

Comments Accepted to CVPR 2022

Lightweight Cross-Lingual Sentence Representation Learning

Zhuoyuan Mao, Prakhar Gupta, Pei Wang, Chenhui Chu, Martin Jaggi, Sadao Kurohashi

专题命中 其他安全 ：alignment(abstract)；分类 cs.CL、cs.AI

Comments ACL 2021 main conference; modified Eq. (2)

Lifelong Personal Context Recognition

Andrea Bontempelli, Marcelo Rodas Britez, Xiaoyue Li, Haonan Zhao, Luca Erculiani, Stefano Teso, Andrea Passerini, Fausto Giunchiglia

专题命中 其他安全 ：alignment(abstract)；分类 cs.AI、cs.LG

Comments 8 pages

Unsupervised Out-of-Domain Detection via Pre-trained Transformers

Keyang Xu, Tongzheng Ren, Shikun Zhang, Yihao Feng, Caiming Xiong

专题命中 其他安全 ：safety(abstract)；分类 cs.CL、cs.LG

Comments Accepted by ACL 2021. Code is available at https://github.com/rivercold/BERT-unsupervised-OOD

Model Agnostic Local Explanations of Reject

André Artelt, Roel Visser, Barbara Hammer

专题命中 其他安全 ：safety(abstract)；分类 cs.AI、cs.LG

Comments arXiv admin note: text overlap with arXiv:2202.07244

Mitigating Toxic Degeneration with Empathetic Data: Exploring the Relationship Between Toxicity and Empathy

Allison Lahnala, Charles Welch, Béla Neuendorf, Lucie Flek

专题命中 其他安全 ：safety(abstract)；分类 cs.CL、cs.AI

Comments Accepted to NAACL 2022

When is BERT Multilingual? Isolating Crucial Ingredients for Cross-lingual Transfer

Ameet Deshpande, Partha Talukdar, Karthik Narasimhan

专题命中 其他安全 ：alignment(abstract)；分类 cs.CL、cs.LG

Comments Accepted at NAACL 2022

Causal Discovery on the Effect of Antipsychotic Drugs on Delirium Patients in the ICU using Large EHR Dataset

Riddhiman Adib, Md Osman Gani, Sheikh Iqbal Ahamed, Mohammad Adibuzzaman

专题命中 其他安全 ：safety(abstract)；分类 cs.AI、cs.LG

Software Testing for Machine Learning

Dusica Marijan, Arnaud Gotlieb

专题命中 其他安全 ：safety(abstract)；分类 cs.AI、cs.LG

Journal ref Proceedings of the AAAI Conference on Artificial Intelligence, 34(09), 13576-13582 (2020)

UBERT: A Novel Language Model for Synonymy Prediction at Scale in the UMLS Metathesaurus

Thilini Wijesiriwardene, Vinh Nguyen, Goonmeet Bajaj, Hong Yung Yip, Vishesh Javangula, Yuqing Mao, Kin Wah Fung, Srinivasan Parthasarathy, Amit P. Sheth, Olivier Bodenreider

专题命中 其他安全 ：alignment(abstract)；分类 cs.CL、cs.AI

Bridging the Gap between Language Models and Cross-Lingual Sequence Labeling

Nuo Chen, Linjun Shou, Ming Gong, Jian Pei, Daxin Jiang

专题命中 其他安全 ：alignment(abstract)；分类 cs.CL、cs.AI

Comments 15 pages

Linguistic communication as (inverse) reward design

Theodore R. Sumers, Robert D. Hawkins, Mark K. Ho, Thomas L. Griffiths, Dylan Hadfield-Menell

专题命中 其他安全 ：alignment(abstract)；分类 cs.CL、cs.AI

Comments 6 pages, 3 figures. Accepted at Learning from Natural Language Supervision workshop (ACL 2022)

MoËT: Mixture of Expert Trees and its Application to Verifiable Reinforcement Learning

Marko Vasic, Andrija Petrovic, Kaiyuan Wang, Mladen Nikolic, Rishabh Singh, Sarfraz Khurshid

专题命中 其他安全 ：safety(abstract)；分类 cs.AI、cs.LG

Journal ref Neural Networks, Volume 151, 2022, Pages 34-47

Aligned Weight Regularizers for Pruning Pretrained Neural Networks

James O' Neill, Sourav Dutta, Haytham Assem

专题命中 其他安全 ：alignment(abstract)；分类 cs.CL、cs.LG

Comments Accepted to ACL Findings 2022

Software Engineering for AI-Based Systems: A Survey

Silverio Martínez-Fernández, Justus Bogner, Xavier Franch, Marc Oriol, Julien Siebert, Adam Trendowicz, Anna Maria Vollmer, Stefan Wagner

专题命中 其他安全 ：safety(abstract)；分类 cs.AI、cs.LG

Comments Accepted in ACM Transactions on Software Engineering and Methodology (TOSEM). For its published version refer to the Journal of ACM TOSEM

Journal ref ACM Trans. Softw. Eng. Methodol. 31, 2, Article 37e (March 2022), 59 pages

Enabling Multimodal Generation on CLIP via Vision-Language Knowledge Distillation

Wenliang Dai, Lu Hou, Lifeng Shang, Xin Jiang, Qun Liu, Pascale Fung

专题命中 其他安全 ：alignment(abstract)；分类 cs.CL、cs.AI

Comments Accepted to ACL 2022

The Ghost in the Machine has an American accent: value conflict in GPT-3

Rebecca L Johnson, Giada Pistilli, Natalia Menédez-González, Leslye Denisse Dias Duran, Enrico Panai, Julija Kalpokiene, Donald Jay Bertulfo

专题命中 其他安全 ：alignment(abstract)；分类 cs.CL、cs.AI

Comments There are a total of 15 pages of the PDF including 8 pages of the main manuscript, 3 pages of references, and 4 pages of appendices. The paper is currently under review by a conference

MLNav: Learning to Safely Navigate on Martian Terrains

Shreyansh Daftry, Neil Abcouwer, Tyler Del Sesto, Siddarth Venkatraman, Jialin Song, Lucas Igel, Amos Byon, Ugo Rosolia, Yisong Yue, Masahiro Ono

专题命中 其他安全 ：safety(abstract)；分类 cs.AI、cs.LG

Comments IEEE Robotics and Automation Letters (RA-L) and ICRA 2022

A Perspective on Robotic Telepresence and Teleoperation using Cognition: Are we there yet?

Hrishav Bakul Barua, Ashis Sau, Ruddra dev Roychoudhury

专题命中 其他安全 ：safety(abstract)；分类 cs.AI、cs.LG

Fanoos: Multi-Resolution, Multi-Strength, Interactive Explanations for Learned Systems

David Bayani, Stefan Mitsch

专题命中 其他安全 ：safety(abstract)；分类 cs.AI、cs.LG

Comments 60 pages, 20 pages main body, 128 references, 3 figures, 5 tables, 12 pseudocode blocks Update 24 Sep. 2020 : Added a pointer to further, external content: Append Section E. Update 20 Mar. 2021: Substantial additions. Further explanations of process, with far more pseudocode. Some corrections to a previous description; see errata section. Also briefly describe a few implemented extensions

The Effects of Reward Misspecification: Mapping and Mitigating Misaligned Models

Alexander Pan, Kush Bhatia, Jacob Steinhardt

专题命中 其他安全 ：safety(abstract)；分类 cs.AI、cs.LG

Comments ICLR 2022; 19 pages

Is High Variance Unavoidable in RL? A Case Study in Continuous Control

Johan Bjorck, Carla P. Gomes, Kilian Q. Weinberger

专题命中 其他安全 ：safety(abstract)；分类 cs.AI、cs.LG

Comments Accepted to ICLR2022

mSLAM: Massively multilingual joint pre-training for speech and text

Ankur Bapna, Colin Cherry, Yu Zhang, Ye Jia, Melvin Johnson, Yong Cheng, Simran Khanuja, Jason Riesa, Alexis Conneau

专题命中 其他安全 ：alignment(abstract)；分类 cs.CL、cs.LG

Roadmap for Cybersecurity in Autonomous Vehicles

Vipin Kumar Kukkala, Sooryaa Vignesh Thiruloga, Sudeep Pasricha

专题命中 其他安全 ：safety(abstract)；分类 cs.AI、cs.LG

A Framework for Pedestrian Sub-classification and Arrival Time Prediction at Signalized Intersection Using Preprocessed Lidar Data

Tengfeng Lin, Zhixiong Jin, Seongjin Choi, Hwasoo Yeo

专题命中 其他安全 ：safety(abstract)；分类 cs.AI、cs.LG

Comments 15 pages, 11 figures, 4 tables

SLAM: A Unified Encoder for Speech and Language Modeling via Speech-Text Joint Pre-Training

Ankur Bapna, Yu-an Chung, Nan Wu, Anmol Gulati, Ye Jia, Jonathan H. Clark, Melvin Johnson, Jason Riesa, Alexis Conneau, Yu Zhang

专题命中 其他安全 ：alignment(abstract)；分类 cs.CL、cs.LG

Single Layer Predictive Normalized Maximum Likelihood for Out-of-Distribution Detection

Koby Bibas, Meir Feder, Tal Hassner

专题命中 其他安全 ：safety(abstract)；分类 cs.AI、cs.LG

Comments NeurIPS 2021

EXplainable Neural-Symbolic Learning (X-NeSyL) methodology to fuse deep learning representations with expert knowledge graphs: the MonuMAI cultural heritage use case

Natalia Díaz-Rodríguez, Alberto Lamas, Jules Sanchez, Gianni Franchi, Ivan Donadello, Siham Tabik, David Filliat, Policarpo Cruz, Rosana Montes, Francisco Herrera

专题命中 其他安全 ：alignment(abstract)；分类 cs.AI、cs.LG

Using Pause Information for More Accurate Entity Recognition

Sahas Dendukuri, Pooja Chitkara, Joel Ruben Antony Moniz, Xiao Yang, Manos Tsagkias, Stephen Pulman

专题命中 其他安全 ：alignment(abstract)；分类 cs.CL、cs.LG

Comments Accepted at the 3rd Workshop on NLP for Conversational AI (EMNLP 2021)