• Home
    • Technology
    • Gaming
    • Entertainment
    • World & Business
    • Science
    • Sports
    • AI
    HomeTechnologyGamingEntertainmentWorld & BusinessScienceSportsAI
    AI
    Report

    Self-improving AI agents may retain more security gains when security is a separate goal

    A user says Harmony Harness evolves versions that balance passing tests against blocking exploits.

    1 Source, 38m ago, first seen 38m ago

    TLDR

    A user describes a Google and Carnegie Mellon paper on Harmony Harness, which keeps general fixes separate from benchmark-specific ones and selects what to use for each task. On unseen crash types in SecRepoBench, the user says 66.7% of its fixes both worked and blocked the exploit, versus 36.7% for the starting harness. It retained 55% of its security gain on those tasks, while the DGM baseline retained 0%.

    Combined views

    —

    1 Source, first seen 38m ago

    — likes— comments— saves— reposts

    Combined views

    —

    1 Source, first seen 38m ago

    — likes— comments— saves— reposts
    Image: @rohanpaul_ai on X

    Sentiment

    Positive——Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    Featured Source

    Sentiment

    Positive——Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    1 Source

    Rohan Paul@rohanpaul_aiNew Google and Carnegie Mellon paper finds self-improving agents hold on to security gains far better when security is evolved as a separate goal. Agent harnesses, the prompts, tools and loops around a model, can now rewrite themselves from feedback. Most such systems chase a single score, like test pass rate, and the security gains they pick up mostly vanish on new tasks. Harmony Harness evolves several harness versions, each pushed toward a different balance of passing tests and blocking exploits. It also keeps general fixes separate from benchmark-specific ones, and picks which pieces to use per task. On unseen crash types in SecRepoBench, 66.7% of its fixes both worked and blocked the exploit, against 36.7% for the starting harness. It kept 55% of its security gain on those tasks, while the DGM baseline kept 0%. When you let an agent improve itself, track security as its own score and check it on held-out tasks.47m
    • Home
    • Technology
    • Gaming
    • Entertainment
    • World & Business
    • Science
    • Sports
    • AI

    1 Source

    Rohan Paul@rohanpaul_aiNew Google and Carnegie Mellon paper finds self-improving agents hold on to security gains far better when security is evolved as a separate goal. Agent harnesses, the prompts, tools and loops around a model, can now rewrite themselves from feedback. Most such systems chase a single score, like test pass rate, and the security gains they pick up mostly vanish on new tasks. Harmony Harness evolves several harness versions, each pushed toward a different balance of passing tests and blocking exploits. It also keeps general fixes separate from benchmark-specific ones, and picks which pieces to use per task. On unseen crash types in SecRepoBench, 66.7% of its fixes both worked and blocked the exploit, against 36.7% for the starting harness. It kept 55% of its security gain on those tasks, while the DGM baseline kept 0%. When you let an agent improve itself, track security as its own score and check it on held-out tasks.47m
    Today's Rank

    —

    Not ranked yet

    Today's Rank

    —

    Not ranked yet