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

    DuoDiscover is designed to find code optimization targets and cheap proxy evaluators together

    A Google team member claims the agent ranked #1 on NVIDIA's SOL-ExecBench and reached 2× human experts on SWE-fficiency.

    Nanyun (Violet) PengN(
    CLSCL
    Dilara SoyluDS
    7 Sources, 2h ago, first seen 2h ago

    TLDR

    A Google team member describes DuoDiscover as a full-stack code optimization agent that co-discovers an optimization target and a cheap proxy evaluator, or “yardstick.” They claim it ranked #1 on NVIDIA's SOL-ExecBench, reached 2× human experts on SWE-fficiency, posted 1.92× on vLLM and 1.44× on FreeToken, and achieved 2.68× over production TPU kernels. They also say it beats the best human- and AI-designed CPU prefetchers.

    Combined views

    7.7K

    7 Sources, first seen 2h ago

    109 likes5 comments34 saves52 reposts

    Combined views

    7.7K

    7 Sources, first seen 2h ago

    109 likes5 comments34 saves52 reposts
    Image: @dilarafsoylu 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.

    7 Sources

    Dilara Soylu@dilarafsoyluAI is cracking math problems. What's holding it back in engineering? The cost of evaluation. Agents can generate ideas in seconds, but testing them can take hours or days. 💡 Meet DuoDiscover: where evaluation is part of what agents discover 👇🧵 Some highlights: 🏆 #1 on NVIDIA's SOL-ExecBench 🐍 2× the experts on SWE-fficiency 🚀 1.92× on vLLM; 1.44× on FreeToken 🔥 2.68× over production TPU kernels 🖥️ Beats the best human and AI-designed CPU prefetchers2h
    Shubham Agarwal@shagarw21🚀 Excited to introduce 𝐃𝐮𝐨𝐃𝐢𝐬𝐜𝐨𝐯𝐞𝐫! This isn’t benchmark gaming. It’s AI agents shipping real PRs and beating human-expert hardware baselines. 🔥 𝐓𝐡𝐞 𝐛𝐫𝐞𝐚𝐤𝐭𝐡𝐫𝐨𝐮𝐠𝐡𝐬 🚀 🛠️ Real optimizations merged into main branches (e.g., FreeToken #619) 🧠 First superhuman CPU prefetcher: +3.93% | +6.31% overall, outperforming AlphaEvolve ⚡ 2.68× TPU kernel speedup | 1.92× faster vLLM serving | 🏆 #1 on NVIDIA SOL-ExecBench (B200) 💡 𝐓𝐡𝐞 𝐬𝐞𝐜𝐫𝐞𝐭 𝐬𝐚𝐮𝐜𝐞? Agents co-discover fast evaluation proxies (“yardsticks”) to slash slow eval bottlenecks and compound real performance gains at scale. 🤝 Incredible collaboration across @GoogleCloud, @BerkeleySky & @StanfordAILab! 🔥 We’re pushing the frontier of AI-driven systems optimization. And we’re just getting started. 🚀2h
    Nanyun (Violet) Peng@VioletNPengSuper excited to share this work from our amazing interns and team at @Google, co-led with awesome @chunliang_tw! Meet DuoDiscover: a full-stack code optimization agent. Its secret sauce is co-discovering the optimization target AND a cheap proxy evaluator (a "yardstick") at the same time. A quick spoiler of the results: 🏆 #1 on NVIDIA's SOL-ExecBench 🐍 2× the human experts on SWE-fficiency 🚀 1.92× on vLLM; 1.44× on FreeToken 🔥 2.68× over production TPU kernels 🖥️ Beats the best human and AI-designed CPU prefetchers Optimizing AI infra. Accelerating science. Dive into more details below 👇2h
    CLS@ChengleiSiRT @dilarafsoylu: AI is cracking math problems. What's holding it back in engineering? The cost of evaluation. Agents can generate ideas i…1h
    Jeff Ma@18jeffreymaso exciting to see our work SWE-fficiency used and some really incredible results on designing agents for codebase optimization! one step closer to self-maintaining codebases :)1h
    Connor Shorten@CShorten30RT @shagarw21: 🚀 Excited to introduce 𝐃𝐮𝐨𝐃𝐢𝐬𝐜𝐨𝐯𝐞𝐫! This isn’t benchmark gaming. It’s AI agents shipping real PRs and beating human-expert…30m
    • Home
    • Technology
    • Gaming
    • Entertainment
    • World & Business
    • Science
    • Sports
    • AI

    7 Sources

    Dilara Soylu@dilarafsoyluAI is cracking math problems. What's holding it back in engineering? The cost of evaluation. Agents can generate ideas in seconds, but testing them can take hours or days. 💡 Meet DuoDiscover: where evaluation is part of what agents discover 👇🧵 Some highlights: 🏆 #1 on NVIDIA's SOL-ExecBench 🐍 2× the experts on SWE-fficiency 🚀 1.92× on vLLM; 1.44× on FreeToken 🔥 2.68× over production TPU kernels 🖥️ Beats the best human and AI-designed CPU prefetchers2h
    Shubham Agarwal@shagarw21🚀 Excited to introduce 𝐃𝐮𝐨𝐃𝐢𝐬𝐜𝐨𝐯𝐞𝐫! This isn’t benchmark gaming. It’s AI agents shipping real PRs and beating human-expert hardware baselines. 🔥 𝐓𝐡𝐞 𝐛𝐫𝐞𝐚𝐤𝐭𝐡𝐫𝐨𝐮𝐠𝐡𝐬 🚀 🛠️ Real optimizations merged into main branches (e.g., FreeToken #619) 🧠 First superhuman CPU prefetcher: +3.93% | +6.31% overall, outperforming AlphaEvolve ⚡ 2.68× TPU kernel speedup | 1.92× faster vLLM serving | 🏆 #1 on NVIDIA SOL-ExecBench (B200) 💡 𝐓𝐡𝐞 𝐬𝐞𝐜𝐫𝐞𝐭 𝐬𝐚𝐮𝐜𝐞? Agents co-discover fast evaluation proxies (“yardsticks”) to slash slow eval bottlenecks and compound real performance gains at scale. 🤝 Incredible collaboration across @GoogleCloud, @BerkeleySky & @StanfordAILab! 🔥 We’re pushing the frontier of AI-driven systems optimization. And we’re just getting started. 🚀2h
    Nanyun (Violet) Peng@VioletNPengSuper excited to share this work from our amazing interns and team at @Google, co-led with awesome @chunliang_tw! Meet DuoDiscover: a full-stack code optimization agent. Its secret sauce is co-discovering the optimization target AND a cheap proxy evaluator (a "yardstick") at the same time. A quick spoiler of the results: 🏆 #1 on NVIDIA's SOL-ExecBench 🐍 2× the human experts on SWE-fficiency 🚀 1.92× on vLLM; 1.44× on FreeToken 🔥 2.68× over production TPU kernels 🖥️ Beats the best human and AI-designed CPU prefetchers Optimizing AI infra. Accelerating science. Dive into more details below 👇2h
    CLS@ChengleiSiRT @dilarafsoylu: AI is cracking math problems. What's holding it back in engineering? The cost of evaluation. Agents can generate ideas i…1h
    Jeff Ma@18jeffreymaso exciting to see our work SWE-fficiency used and some really incredible results on designing agents for codebase optimization! one step closer to self-maintaining codebases :)1h
    Connor Shorten@CShorten30RT @shagarw21: 🚀 Excited to introduce 𝐃𝐮𝐨𝐃𝐢𝐬𝐜𝐨𝐯𝐞𝐫! This isn’t benchmark gaming. It’s AI agents shipping real PRs and beating human-expert…30m
    Today's Rank

    #13

    Today's Rank

    #13