본문

서브메뉴

Promises and Pitfalls of Generative AI: An AI-Safety Centric Approach- [electronic resource]
Promises and Pitfalls of Generative AI: An AI-Safety Centric Approach - [electronic resour...
Promises and Pitfalls of Generative AI: An AI-Safety Centric Approach- [electronic resource]

상세정보

자료유형  
 학위논문파일 국외
최종처리일시  
20240214100446
ISBN  
9798379717469
DDC  
621.3
저자명  
Sehwag, Vikash.
서명/저자  
Promises and Pitfalls of Generative AI: An AI-Safety Centric Approach - [electronic resource]
발행사항  
[S.l.]: : Princeton University., 2023
발행사항  
Ann Arbor : : ProQuest Dissertations & Theses,, 2023
형태사항  
1 online resource(171 p.)
주기사항  
Source: Dissertations Abstracts International, Volume: 84-12, Section: B.
주기사항  
Advisor: Mittal, Prateek;Chiang, Mung.
학위논문주기  
Thesis (Ph.D.)--Princeton University, 2023.
사용제한주기  
This item must not be sold to any third party vendors.
초록/해제  
요약Artificial intelligence (AI) has advanced rapidly, leading to remarkable progress across numerous real-world applications. However, the prevalence of AI-enabled decisions also raises concerns about its potential safety risks, as AI systems are known to exhibit failure cases across multiple domains, such as autonomous driving, medical diagnostics, and content moderation. In this thesis, we investigate AI safety challenges through the lens of generative models, a class of machine learning models capable of approximating the underlying distribution of training datasets and synthesizing novel samples. By bridging the gap between generative models and AI safety, we reveal the immense potential of generative models in addressing safety challenges, while also identifying safety risks posed by contemporary generative models.First, we focus on improving generalization in adversarially robust learning with generative models by incorporating them into existing machine learning pipelines and distilling their knowledge by synthesizing novel synthetic images. We assess various generative models and propose a new metric (ARC), based on the indistinguishability of adversarially perturbed synthetic and real data, to accurately determine the generalization benefit of different generative models. Next, we investigate task-aware knowledge distillation from generative models, where we first demonstrate the disparate contributions of individual synthetic images in improving generalization. To adaptively sample images with the highest generalization benefit, we propose an adaptive sampling technique that guides the sampling process in diffusion models to maximize the generalization benefit of generated synthetic images.Next, we address the shortcomings of long-tailed data distributions, which underlie numerous challenges in AI safety, by using generative models to generate high-fidelity samples from low-density regions. We propose a novel low-density sampling process for diffusion models, guiding the process towards low-density regions while maintaining fidelity, and rigorously demonstrate that our process successfully generates novel high-fidelity samples from low-density regions.Finally, we demonstrate some of the key limitations of existing generative models. We first consider the outlier detection task and demonstrate the shortcomings of modern generative models in solving it. Considering our findings, we propose SSD, an unsupervised framework for outlier detection based on unlabeled in-distribution data. We further uncover that modern diffusion models, which are used by millions of users, leak the privacy of training data, where we extract a nontrivial number of training images from the pre-trained diffusion models.In summary, this thesis addresses multiple AI safety challenges and provides a comprehensive framework for the safety and reliability of AI systems under the new generative AI paradigm.
일반주제명  
Computer engineering.
일반주제명  
Electrical engineering.
일반주제명  
Computer science.
키워드  
Adversarial robustness
키워드  
Deep neural networks
키워드  
Generative artificial intelligence
키워드  
Privacy risks
키워드  
Trustworthy machine learning
기타저자  
Princeton University Electrical and Computer Engineering
기본자료저록  
Dissertations Abstracts International. 84-12B.
기본자료저록  
Dissertation Abstract International
전자적 위치 및 접속  
로그인 후 원문을 볼 수 있습니다.

MARC

 008240612s2023      us  |||||||||||||||c||eng  d
■001000016932350
■00520240214100446
■006m          o    d                
■007cr#unu||||||||
■020    ▼a9798379717469
■035    ▼a(MiAaPQ)AAI30491778
■040    ▼aMiAaPQ▼cMiAaPQ
■0820  ▼a621.3
■1001  ▼aSehwag,  Vikash.
■24510▼aPromises  and  Pitfalls  of  Generative  AI:  An  AI-Safety  Centric  Approach▼h[electronic  resource]
■260    ▼a[S.l.]:▼bPrinceton  University.  ▼c2023
■260  1▼aAnn  Arbor  :▼bProQuest  Dissertations  &  Theses,  ▼c2023
■300    ▼a1  online  resource(171  p.)
■500    ▼aSource:  Dissertations  Abstracts  International,  Volume:  84-12,  Section:  B.
■500    ▼aAdvisor:  Mittal,  Prateek;Chiang,  Mung.
■5021  ▼aThesis  (Ph.D.)--Princeton  University,  2023.
■506    ▼aThis  item  must  not  be  sold  to  any  third  party  vendors.
■520    ▼aArtificial  intelligence  (AI)  has  advanced  rapidly,  leading  to  remarkable  progress  across  numerous  real-world  applications.  However,  the  prevalence  of  AI-enabled  decisions  also  raises  concerns  about  its  potential  safety  risks,  as  AI  systems  are  known  to  exhibit  failure  cases  across  multiple  domains,  such  as  autonomous  driving,  medical  diagnostics,  and  content  moderation.  In  this  thesis,  we  investigate  AI  safety  challenges  through  the  lens  of  generative  models,  a  class  of  machine  learning  models  capable  of  approximating  the  underlying  distribution  of  training  datasets  and  synthesizing  novel  samples.  By  bridging  the  gap  between  generative  models  and  AI  safety,  we  reveal  the  immense  potential  of  generative  models  in  addressing  safety  challenges,  while  also  identifying  safety  risks  posed  by  contemporary  generative  models.First,  we  focus  on  improving  generalization  in  adversarially  robust  learning  with  generative  models  by  incorporating  them  into  existing  machine  learning  pipelines  and  distilling  their  knowledge  by  synthesizing  novel  synthetic  images.  We  assess  various  generative  models  and  propose  a  new  metric  (ARC),  based  on  the  indistinguishability  of  adversarially  perturbed  synthetic  and  real  data,  to  accurately  determine  the  generalization  benefit  of  different  generative  models.  Next,  we  investigate  task-aware  knowledge  distillation  from  generative  models,  where  we  first  demonstrate  the  disparate  contributions  of  individual  synthetic  images  in  improving  generalization.  To  adaptively  sample  images  with  the  highest  generalization  benefit,  we  propose  an  adaptive  sampling  technique  that  guides  the  sampling  process  in  diffusion  models  to  maximize  the  generalization  benefit  of  generated  synthetic  images.Next,  we  address  the  shortcomings  of  long-tailed  data  distributions,  which  underlie  numerous  challenges  in  AI  safety,  by  using  generative  models  to  generate  high-fidelity  samples  from  low-density  regions.  We  propose  a  novel  low-density  sampling  process  for  diffusion  models,  guiding  the  process  towards  low-density  regions  while  maintaining  fidelity,  and  rigorously  demonstrate  that  our  process  successfully  generates  novel  high-fidelity  samples  from  low-density  regions.Finally,  we  demonstrate  some  of  the  key  limitations  of  existing  generative  models.  We  first  consider  the  outlier  detection  task  and  demonstrate  the  shortcomings  of  modern  generative  models  in  solving  it.  Considering  our  findings,  we  propose  SSD,  an  unsupervised  framework  for  outlier  detection  based  on  unlabeled  in-distribution  data.  We  further  uncover  that  modern  diffusion  models,  which  are  used  by  millions  of  users,  leak  the  privacy  of  training  data,  where  we  extract  a  nontrivial  number  of  training  images  from  the  pre-trained  diffusion  models.In  summary,  this  thesis  addresses  multiple  AI  safety  challenges  and  provides  a  comprehensive  framework  for  the  safety  and  reliability  of  AI  systems  under  the  new  generative  AI  paradigm.
■590    ▼aSchool  code:  0181.
■650  4▼aComputer  engineering.
■650  4▼aElectrical  engineering.
■650  4▼aComputer  science.
■653    ▼aAdversarial  robustness
■653    ▼aDeep  neural  networks
■653    ▼aGenerative  artificial  intelligence
■653    ▼aPrivacy  risks
■653    ▼aTrustworthy  machine  learning
■690    ▼a0464
■690    ▼a0544
■690    ▼a0984
■690    ▼a0800
■71020▼aPrinceton  University▼bElectrical  and  Computer  Engineering.
■7730  ▼tDissertations  Abstracts  International▼g84-12B.
■773    ▼tDissertation  Abstract  International
■790    ▼a0181
■791    ▼aPh.D.
■792    ▼a2023
■793    ▼aEnglish
■85640▼uhttp://www.riss.kr/pdu/ddodLink.do?id=T16932350▼nKERIS▼z이  자료의  원문은  한국교육학술정보원에서  제공합니다.
■980    ▼a202402▼f2024

미리보기

내보내기

chatGPT토론

Ai 추천 관련 도서


    신착도서 더보기
    최근 3년간 통계입니다.

    소장정보

    • 예약
    • 소재불명신고
    • 나의폴더
    • 우선정리요청
    • 비도서대출신청
    • 야간 도서대출신청
    소장자료
    등록번호 청구기호 소장처 대출가능여부 대출정보
    TF07266 전자도서 마이폴더 부재도서신고 비도서대출신청

    * 대출중인 자료에 한하여 예약이 가능합니다. 예약을 원하시면 예약버튼을 클릭하십시오.

    해당 도서를 다른 이용자가 함께 대출한 도서

    관련 인기도서

    로그인 후 이용 가능합니다.