Close Menu
Ztoog
    What's Hot
    Mobile

    Motorola Razr 40 and Razr 40 Ultra go global, will be called Razr and Razr+ in North America

    AI

    Meet Plandex: An Open-Source Terminal-based AI Coding Engine for Complex Tasks

    AI

    In a New AI Paper, CMU and Google Researchers Redefine Language Model Outputs: How Delaying Responses with Pause Tokens Boosts Performance on QA and Reasoning Tasks

    Important Pages:
    • About Us
    • Contact us
    • Privacy Policy
    • Terms & Conditions
    Facebook X (Twitter) Instagram Pinterest
    Facebook X (Twitter) Instagram Pinterest
    Ztoog
    • Home
    • The Future

      JD Vance and President Trump’s Sons Hype Bitcoin at Las Vegas Conference

      AI may already be shrinking entry-level jobs in tech, new research suggests

      Today’s NYT Strands Hints, Answer and Help for May 26 #449

      LiberNovo Omni: The World’s First Dynamic Ergonomic Chair

      Common Security Mistakes Made By Businesses and How to Avoid Them

    • Technology

      Gemini in Google Drive can now help you skip watching that painfully long Zoom meeting

      Apple iPhone exports from China to the US fall 76% as India output surges

      Today’s NYT Wordle Hints, Answer and Help for May 26, #1437

      5 Skills Kids (and Adults) Need in an AI World – O’Reilly

      How To Come Back After A Layoff

    • Gadgets

      8 Best Vegan Meal Delivery Services and Kits (2025), Tested and Reviewed

      Google Home is getting deeper Gemini integration and a new widget

      Google Announces AI Ultra Subscription Plan With Premium Features

      Google shows off Android XR-based glasses, announces Warby Parker team-up

      The market’s down, but this OpenAI for the stock market can help you trade up

    • Mobile

      Microsoft is done being subtle – this new tool screams “upgrade now”

      Wallpaper Wednesday: Android wallpapers 2025-05-28

      Google can make smart glasses accessible with Warby Parker, Gentle Monster deals

      vivo T4 Ultra specs leak

      Forget screens: more details emerge on the mysterious Jony Ive + OpenAI device

    • Science

      Do we have free will? Quantum experiments may soon reveal the answer

      Was Planet Nine exiled from the solar system as a baby?

      How farmers can help rescue water-loving birds

      A trip to the farm where loofahs grow on vines

      AI Is Eating Data Center Power Demand—and It’s Only Getting Worse

    • AI

      The AI Hype Index: College students are hooked on ChatGPT

      Learning how to predict rare kinds of failures | Ztoog

      Anthropic’s new hybrid AI model can work on tasks autonomously for hours at a time

      AI learns how vision and sound are connected, without human intervention | Ztoog

      How AI is introducing errors into courtrooms

    • Crypto

      GameStop bought $500 million of bitcoin

      CoinW Teams Up with Superteam Europe to Conclude Solana Hackathon and Accelerate Web3 Innovation in Europe

      Ethereum Net Flows Turn Negative As Bulls Push For $3,500

      Bitcoin’s Power Compared To Nuclear Reactor By Brazilian Business Leader

      Senate advances GENIUS Act after cloture vote passes

    Ztoog
    Home » IBM Researchers Propose a New Adversarial Attack Framework Capable of Generating Adversarial Inputs for AI Systems Regardless of the Modality or Task
    AI

    IBM Researchers Propose a New Adversarial Attack Framework Capable of Generating Adversarial Inputs for AI Systems Regardless of the Modality or Task

    Facebook Twitter Pinterest WhatsApp
    IBM Researchers Propose a New Adversarial Attack Framework Capable of Generating Adversarial Inputs for AI Systems Regardless of the Modality or Task
    Share
    Facebook Twitter LinkedIn Pinterest WhatsApp

    In the ever-evolving panorama of synthetic intelligence, a rising concern has emerged. The vulnerability of AI fashions to adversarial evasion assaults. These crafty exploits can result in deceptive mannequin outputs with refined alterations in enter information, a risk extending past pc imaginative and prescient fashions. The want for strong defenses towards such assaults is obvious as AI deeply integrates into our each day lives.

    Due to their numerical nature, present efforts to fight adversarial assaults have primarily targeted on pictures, making them handy targets for manipulation. While substantial progress has been made on this area, different information varieties, akin to textual content and tabular information, current distinctive challenges. These information varieties have to be remodeled into numerical function vectors for mannequin consumption, and their semantic guidelines have to be preserved throughout adversarial modifications. Most out there toolkits need assistance to deal with these complexities, leaving AI fashions in these domains susceptible.

    URET is a game-changer in the battle towards adversarial assaults. URET treats malicious assaults as a graph exploration drawback, with every node representing an enter state and every edge representing an enter transformation. It effectively identifies sequences of modifications that result in mannequin misclassification. The toolkit provides a easy configuration file on GitHub, permitting customers to outline exploration strategies, transformation varieties, semantic guidelines, and aims tailor-made to their wants.

    In a current paper from IBM analysis, the URET group demonstrated its prowess by producing adversarial examples for tabular, textual content, and file enter varieties, all supported by URET’s transformation definitions. However, URET’s true power lies in its flexibility. Recognizing the huge range of machine studying implementations, the toolkit gives an open door for superior customers to outline personalized transformations, semantic guidelines, and exploration aims. 

    URET depends on metrics highlighting its effectiveness in producing adversarial examples throughout numerous information varieties to measure its capabilities. These metrics reveal URET’s skill to establish and exploit vulnerabilities in AI fashions whereas additionally offering a standardized means of evaluating mannequin robustness towards evasion assaults.

    In conclusion, the introduction of AI has ushered in a new period of innovation, however it has additionally introduced forth new challenges, akin to adversarial evasion assaults. The Universal Robustness Evaluation Toolkit (URET) for evasion emerges as a beacon of hope on this evolving panorama. With its graph exploration method, adaptability to totally different information varieties, and a rising group of open-source contributors, URET represents a important step towards safeguarding AI techniques from malicious threats. As machine studying continues to permeate numerous features of our lives, the rigorous analysis and evaluation supplied by URET stand as the greatest protection towards adversarial vulnerabilities, guaranteeing the continued trustworthiness of AI in our more and more interconnected world.


    Check out the Paper, GitHub hyperlink, and Reference Article. All Credit For This Research Goes To the Researchers on This Project. Also, don’t neglect to affix our 30k+ ML SubReddit, 40k+ Facebook Community, Discord Channel, and Email Newsletter, the place we share the newest AI analysis information, cool AI tasks, and extra.

    If you want our work, you’ll love our e-newsletter..


    Niharika is a Technical consulting intern at Marktechpost. She is a third 12 months undergraduate, at present pursuing her B.Tech from Indian Institute of Technology(IIT), Kharagpur. She is a extremely enthusiastic particular person with a eager curiosity in Machine studying, Data science and AI and an avid reader of the newest developments in these fields.


    🚀 The finish of challenge administration by people (Sponsored)

    Share. Facebook Twitter Pinterest LinkedIn WhatsApp

    Related Posts

    AI

    The AI Hype Index: College students are hooked on ChatGPT

    AI

    Learning how to predict rare kinds of failures | Ztoog

    AI

    Anthropic’s new hybrid AI model can work on tasks autonomously for hours at a time

    AI

    AI learns how vision and sound are connected, without human intervention | Ztoog

    AI

    How AI is introducing errors into courtrooms

    AI

    With AI, researchers predict the location of virtually any protein within a human cell | Ztoog

    AI

    Google DeepMind’s new AI agent cracks real-world problems better than humans can

    AI

    Study shows vision-language models can’t handle queries with negation words | Ztoog

    Leave A Reply Cancel Reply

    Follow Us
    • Facebook
    • Twitter
    • Pinterest
    • Instagram
    Top Posts
    The Future

    The Future of iGaming Lies in the Hands of AI

    In a quickly evolving digital world, industries are compelled to undertake modern applied sciences to…

    AI

    DenseFormer by EPFL Researchers: Enhancing Transformer Efficiency with Depth-Weighted Averages for Superior Language Modeling Performance and Speed

    The transformer structure has improved pure language processing, with latest developments achieved by scaling efforts…

    Crypto

    Finance Guru Reveals Why Bitcoin Is The ‘Perfect Asset At The Right Time’

    Robert Kiyosaki, the best-selling creator of the favored guide, ‘Rich Dad Poor Dad’ has hailed…

    Mobile

    Google preemptively marks Pixel’s scrolling stutter issue as fixed on Android 15

    Google Pixel customers ought to put together for a probably smoother expertise starting with Android…

    Science

    Sweat Is Helping You Survive Climate Change

    The advertising marketing campaign was a long-lasting success, even a century later. Last 12 months,…

    Our Picks
    The Future

    ‘John Wick: Chapter 4’ Streaming Release Date and How to Watch From Anywhere

    Mobile

    This amazing Samsung Galaxy Tab S8 Ultra deal ends tonight

    Gadgets

    48 Best Back-to-School Deals (2023): Laptops, Backpacks, Household Essentials

    Categories
    • AI (1,492)
    • Crypto (1,753)
    • Gadgets (1,804)
    • Mobile (1,850)
    • Science (1,865)
    • Technology (1,801)
    • The Future (1,647)
    Most Popular
    Technology

    Experts detail how users are getting scammed on Facebook Marketplace; Meta says it plans a notification system to let users identify "scams around payment apps" (Amanda Hoover/Wired)

    Crypto

    Resy and Eater co-founder raises $24M for Blackbird, a restaurant loyalty platform

    Technology

    Testing the 2024 BMW M2—maybe the last M car with a manual transmission

    Ztoog
    Facebook X (Twitter) Instagram Pinterest
    • Home
    • About Us
    • Contact us
    • Privacy Policy
    • Terms & Conditions
    © 2025 Ztoog.

    Type above and press Enter to search. Press Esc to cancel.