与「AutoML」相关的搜索结果

开源项目 LoopX：超长程 Agent 自主运行 200+ hours，状态不漂移。 我的技术主张是：LLM 上下文有限，长程 Agent 需要外置状态，通过完备的状态管理、监督和规划，让 Agent 无人干预时跑得稳、持续有产出；有人干预时跑得更好，能吸收反馈继续演进。 两条真实 trajectory 分别跨越 220.7 / 272.9 小时，跨多轮执行、等待、人工决策、writeback 与 resume 后，整个 loop 仍能找回目标、证据和下一步。 目前 LoopX 已有 3 个 showcase：auto PR issue fix、AutoML experiment 和 auto coredump fix。 以 OpenViking 开源仓库的 PR issue fix 为例，Agent 不只是循环写代码。它需要持续理解 issue 的不同状态，判断何时开发、何时等待、何时请求 review，处理 CI、冲突和上游变化，并连续交付多个 PR。 这对应 LoopX 的 domain state 管理：领域系统决定真实状态，LoopX 负责把状态投影成下一步可执行的工作。 与此同时，Agent 还可以在干活过程中实现能力自进化。当它发现现有系统缺少某项能力时，可以提出 feature、完成开发与验证、发布新的离线或在线版本，再使用新能力继续原来的任务。 长程 Agent 天然适合自进化，“完成工作”和“升级完成工作的系统”可以在同一条长程轨迹中发生。 LoopX 把这些信息外置成结构化控制面： • Goal / Vision：目标是什么，什么不能被局部优化牺牲 • Todo / Gate：当前执行的 frontier，以及必须留给人的关键判断 • Identity / Authority：谁能 claim、writeback、approve • Evidence / Receipt：每次推进留下什么可回读证据 • Quota / Scheduler：何时继续执行，何时安静等待 • Handoff / Recovery：换模型、换会话、换 host 后如何恢复 你也可以把它理解为一块专门给 Agent 设计的可执行 Kanban。 普通看板只展示“谁在做什么”；LoopX 的状态会直接约束和驱动下一次 bounded turn，让看板本身成为执行系统的一部分。 这套系统最强的地方是通用性。它不只可以修 PR，还可以做 auto research、长期实验、自媒体运营、复杂 feature 开发和办公任务。 Agent 不再只是一次性的回答机器，而可以围绕一个人的 vision，长期工作、等待、吸收反馈、积累证据并持续演进。 LoopX 从一开始 build in public：状态协议、CLI、控制面实现和真实运行轨迹都进入了开源仓库；它也已经和 OpenViking、NoKV 等 agent infra 项目形成了开源合作伙伴关系。 我希望 LoopX 最终能放大每个人的 vision 和想象力。只要你有自己的目标、想法或技术主张，就能拥有一个全天候继续工作和探索的 Agent 系统，帮助你把愿景一点点变成现实。 欢迎试用、提 issue、贡献代码，或者用一个真实的 multi-day task 跑 LoopX。

🧵 Deli AutoResearch SKILL is now officially open source! 🎉 Alongside it, we’re dropping our 4th survey paper — this time on Self-play. Inspired by AlphaZero, we got a powerful insight: prior knowledge doesn’t always lift the ceiling. Models can discover more globally optimal solutions just by playing against themselves. The biggest change in this paper? For the first time, the AutoResearch Agent autonomously planned GPU experiments — and submitted actual RL runs on the DeepSeek 285B model. The entire RL pipeline — experiment design, code writing, running, debugging, and conclusion summarization — was 100% automated, with zero human intervention from me. This was incredibly difficult, but an incredibly important step. GRPO is the tool being called by the AutoResearch Agent here. We see this as the beginning of our Continual Learning research journey. 🚀 As always, this is my personal research project, unaffiliated with any organization. All views are my own. #AI# #ReinforcementLearning# #SelfPlay# #OpenSource# #AutoML# #ContinualLearning# #DeepSeek#

If you're curious about AutoML, check out this book, co-authored by @haifeng_jin from the Keras team. You'll learn about how you can automate your ML pipelines with AutoKeras and KerasTuner!

25,000,000 $COMBULL - roughly $1,500 at today’s price. Locked until Halloween. Split evenly between every recipient wallet. No tricks. Just treats. How to get in: → Hold $50+ in $COMBULL → Show up for the community - X, Discord, TikTok, anywhere the herd is → Submit your wallet No whitelist. No lottery. No inner circle. Every wallet goes into the contract. October 31 it unlocks and pays out automatically. The pool is fixed. The price isn’t. Hold. Post. Get counted 🎃 Web:

Gemini 3.7 Flash is live! - Low latency and lower token costs by 50% while advancing reasoning for coding and agentic loops. - Software Engineering Performance (DeepSWE v1.1): 37.0% ➔ 65.3% - Enterprise Automation (AutomationBench): 13.4% ➔ 30.4% Available now via our APIs, @GoogleAIStudio and @antigravity!

Say hello to Gemini 3.7 Flash, generally available today! ⚡️We took your feedback and pushed hard on real-world coding benchmarks and agent autonomy: - Major jumps on coding and agentic use cases - DeepSWE (65.3% vs 49.0%) FrontierCode (43.6% vs 34.4%), AutomationBench (30.4% vs 17.0%) - 50% discount until EOY: $0.75/$3.75 per 1M tokens - Default model in Managed Agents, @GoogleAIStudio , @antigravity, and @GeminiApp Spark Throw your hardest coding/agentic tasks at it and let us know what you build:

I’ve had my @Tesla for 9 days—and it’s already saved my life. Full Self-Driving with HW4 is mind-blowing. My car drove me ~800 miles from Georgia to Texas without me touching the wheel or pedals, while avoiding multiple collisions. I’ve never felt safer in a vehicle. Here’s one example of a collision avoided, automatically captured on camera. In this instance (though it doesn’t show my actual POV), sparks covered the windshield, lighting up the road like the Fourth of July and obstructing my view. Since I was in FSD, the car reacted immediately without overcorrecting like I likely would have, which could have led to a far worse outcome. Everyone who’s ridden with me and experienced FSD for the first time is blown away and now looking into buying one—including my parents, who like me have never owned an EV. There’s no telling how many lives this technology will save. This is hands-down the most incredible purchase I’ve ever made. God bless @elonmusk & the Tesla team🫡