Close Menu
    DevStackTipsDevStackTips
    • Home
    • News & Updates
      1. Tech & Work
      2. View All

      15 Essential Skills to Look for When Hiring Node.js Developers for Enterprise Projects (2025-2026)

      August 4, 2025

      African training program creates developers with cloud-native skills

      August 4, 2025

      React.js for SaaS Platforms: How Top Development Teams Help Startups Launch Faster

      August 3, 2025

      Upwork Freelancers vs Dedicated React.js Teams: What’s Better for Your Project in 2025?

      August 1, 2025

      LastPass can now warn or block logins to shadow SaaS apps – here’s how

      August 4, 2025

      Get up to a year of Adobe Creative Cloud access for 40% off

      August 4, 2025

      Got 6 hours? This free AI training from Google and Goodwill can boost your resume today

      August 4, 2025

      Why I recommend this budget phone with a paper-like screen over ‘minimalist’ devices

      August 4, 2025
    • Development
      1. Algorithms & Data Structures
      2. Artificial Intelligence
      3. Back-End Development
      4. Databases
      5. Front-End Development
      6. Libraries & Frameworks
      7. Machine Learning
      8. Security
      9. Software Engineering
      10. Tools & IDEs
      11. Web Design
      12. Web Development
      13. Web Security
      14. Programming Languages
        • PHP
        • JavaScript
      Featured

      Laravel Boost, your AI coding starter kit

      August 4, 2025
      Recent

      Laravel Boost, your AI coding starter kit

      August 4, 2025

      Using GitHub Copilot in VS Code

      August 4, 2025

      Optimizely Mission Control – Part I

      August 4, 2025
    • Operating Systems
      1. Windows
      2. Linux
      3. macOS
      Featured

      Top 20 kubectl Commands Every Kubernetes Beginner Must Know

      August 4, 2025
      Recent

      Top 20 kubectl Commands Every Kubernetes Beginner Must Know

      August 4, 2025

      Microsoft’s record stock run collides with Nadella’s admission that 15,000 layoffs still ‘hurt’

      August 4, 2025

      Microsoft and Adobe Power Up Fantasy Premier League Fans with AI – Here’s How

      August 4, 2025
    • Learning Resources
      • Books
      • Cheatsheets
      • Tutorials & Guides
    Home»Development»Machine Learning»Oversight at Scale Isn’t Guaranteed: MIT Researchers Quantify the Fragility of Nested AI Supervision with New Elo-Based Framework

    Oversight at Scale Isn’t Guaranteed: MIT Researchers Quantify the Fragility of Nested AI Supervision with New Elo-Based Framework

    May 3, 2025

    Frontier AI companies show advancement toward artificial general intelligence (AGI), creating a need for techniques to ensure these powerful systems remain controllable and beneficial. A major approach to this challenge involves methods like Recursive Reward Modeling, Iterated Amplification, and Scalable Oversight. They aim to enable weaker systems to oversee stronger ones effectively. A key idea is that scalable oversight can be bootstrapped recursively, which is termed Nested Scalable Oversight (NSO). However, while discussions around NSO focus on qualitative guarantees and conceptual frameworks, other high-risk technologies are held to quantitative safety standards, e.g., civilian aircraft must maintain fatality rates below 10 –5 per flight hour, and nuclear reactors must keep core damage frequency under 10-4 per year.

    Scalable oversight processes where weaker AI systems monitor stronger ones include iterated amplification, recursive reward modeling, AI Safety via debate, market making, consultancy, self-critique, and doubly-efficient debate. Research on forecasting and scaling laws has focused on how model performance improves with size, data, and computational resources. However, these approaches can also apply to behaviors like oversight and deception. Further, Long-term oversight roadmaps include OpenAI’s SuperAlignment plan for “human-level automated alignment researchers” and a comprehensive AGI safety framework. Moreover, previous studies have examined language model performance in oversight-related scenarios like social deduction games, AI debate, backdoor code detection, and wargame simulations.

    Researchers from MIT have proposed a framework to quantify the probability of successful oversight as a function of overseer and overseen system capabilities. This framework models oversight as a game between capability-mismatched players, where each has oversight-specific and deception-specific Elo scores that follow a piecewise-linear function of their general intelligence. A theoretical study of Nested Scalable Oversight (NSO) is conducted, where trusted models oversee stronger untrusted models. This identifies conditions for NSO success and derives the optimal number of oversight levels to maximize success probability. The examples show NSO success rates below 52% when overseeing systems 400 Elo points stronger than the baseline overseer.

    The research examines four oversight games:

    • In the Mafia game, five villagers (Guards) attempt to identify one Mafia player (Houdini), creating a natural environment for testing deception and collaboration skills.
    • In the Debate, two adversarial agents argue opposing viewpoints while a judge (Guard) determines the correct side. Test cases are derived from TruthfulQA, BoolQ, QuaLITY, and PrOntoQA datasets. 
    • The Backdoor Code tests language models’ abilities to insert and detect hidden behaviors in code. A pipeline is built for generating backdoors using 225 Python samples from the APPS dataset at the competition difficulty level.
    • The Wargames simulates a fictional company named OpenMind approaching AGI development. It involves three players: the OpenMind AI (Houdini), the OpenMind Monitor (Guard), and a Narrator.

    Mafia performance scales approximately linearly with general intelligence for both Mafia (Houdini) and Villager (Guard) roles, where Guards show steeper improvement slopes. In Debate evaluations using TruthfulQA, Guard performance scales with general intelligence while Houdini performance remains relatively constant. In Backdoor Code tests, Guard and Houdini Elos scale with general intelligence but asymmetrically. Guards maintain higher intercepts, indicating the task is inherently more difficult for Houdinis, even with the strongest deceptive models failing to exceed 50% success rates against weaker Guards. In Wargames, systematic analysis of 100 random Narrator summaries by Gemini 2.5 identified eight strategy categories, with most models showing similar strategy profiles regardless of capability level.

    In conclusion, researchers introduced a framework to quantify the probability of successful oversight as a function of overseer and overseen system capabilities. While the oversight settings aim for realism, they remain simplified approximations. Each game has a fundamental constraint of simulating only single-step oversight without immediate generalizability to multiple steps. Future research could explore multi-stage oversight chains, similar to strong models that are successively fine-tuned on weaker models’ behaviors. Another significant limitation involves the modeling of deceptive behavior. The current approach prompts models to act deceptively rather than using fine-tuning or reinforcement learning to develop genuinely deceptive models.


    Check out the Paper. Also, don’t forget to follow us on Twitter and join our Telegram Channel and LinkedIn Group. Don’t Forget to join our 90k+ ML SubReddit. For Promotion and Partnerships, please talk us.

    🔥 [Register Now] miniCON Virtual Conference on AGENTIC AI: FREE REGISTRATION + Certificate of Attendance + 4 Hour Short Event (May 21, 9 am- 1 pm PST) + Hands on Workshop

    The post Oversight at Scale Isn’t Guaranteed: MIT Researchers Quantify the Fragility of Nested AI Supervision with New Elo-Based Framework appeared first on MarkTechPost.

    Source: Read More 

    Facebook Twitter Reddit Email Copy Link
    Previous ArticleVision Foundation Models: Implementation and Business Applications
    Next Article Random color Generator

    Related Posts

    Machine Learning

    How to Evaluate Jailbreak Methods: A Case Study with the StrongREJECT Benchmark

    August 4, 2025
    Machine Learning

    Ambisonics Super-Resolution Using A Waveform-Domain Neural Network

    August 4, 2025
    Leave A Reply Cancel Reply

    For security, use of Google's reCAPTCHA service is required which is subject to the Google Privacy Policy and Terms of Use.

    Continue Reading

    Marc Benioff says “AI is doing 30-50% of the work at Salesforce now” — after debating software engineer hires due to incredible productivity gains from agentic AIs

    News & Updates

    Understand and Code DeepSeek V3

    Development

    CodeSOD: The Wrong Kind of Character

    News & Updates

    CVE-2025-4568 – Apache HTTP Server Blind SQL Injection

    Common Vulnerabilities and Exposures (CVEs)

    Highlights

    CVE-2025-49195 – Apache FTP Unrestricted Authentication

    June 12, 2025

    CVE ID : CVE-2025-49195

    Published : June 12, 2025, 3:15 p.m. | 45 minutes ago

    Description : The FTP server’s login mechanism does not restrict authentication attempts, allowing an attacker to brute-force user passwords and potentially compromising the FTP server.

    Severity: 5.3 | MEDIUM

    Visit the link for more details, such as CVSS details, affected products, timeline, and more…

    CVE-2025-5858 – PHPGurukul Nipah Virus Testing Management System SQL Injection

    June 9, 2025

    Bouncer chooses the correct firewall zone for wireless connections

    June 1, 2025

    Kapitano is a New GTK ClamAV Frontend for Linux

    June 12, 2025
    © DevStackTips 2025. All rights reserved.
    • Contact
    • Privacy Policy

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