ScienceFlow: A long-horizon agent for ML research, scientific discovery and beyond
面向ML研究与科学发现的超长时程智能体,自主推演实验闭环,值得AI从业者关注。
arXiv:2608.14354v1 Announce Type: new Abstract: Enabling LLM agents to sustain productive, stable, and goal-aligned research over extended horizons is…
Fisher-R1: Training LLM Agents for Reliable Hypothesis Testing
Fisher-R1用强化学习训练AI智能体做统计假说检验,显著提升可靠性,值得关注的研究。
arXiv:2608.07437v1 Announce Type: new Abstract: Reliable hypothesis testing is the foundation of many empirical scientific claims. Large language mode…
Accelerating scientific discovery with ChatGPT for Academic Researchers
OpenAI向10万名学术研究者免费开放ChatGPT最强模型,助力科研协作与发现加速。
OpenAI is giving 100,000 academic researchers free access to ChatGPT's most advanced AI models to accelerate scientific research, collaboration, and d…
浙江:支持运用人工智能赋能各领域科学发现与技术创新
浙江省委省政府发布关于建设一流创新生态打造最具竞争力营商环境的意见。意见指出,以人工智能赋能创新范式变革。持续深化“数字浙江”建设,营造最优开源开放生态。全面实施“人工智能+”行动,布局建设科学基础模型,支持运用人工智能赋能各领域科学发现与技术创新,推进人工智能同产业发展、消费提质、文化建设、民生保…
腾讯混元发布科学发现智能体 Hyra,能够递归自我改进
腾讯混元推出递归自我改进的科研智能体Hyra,与DeepMind、Together AI等前沿进展对比,展现AI科学发现新突破
IT之家 7 月 21 日消息,腾讯混元今日推出 Hyra-1.0 ,这是 Hyra(Hunyuan Research Agent)的首个版本:一个能够递归自我改进、专为性能导向的研究与工程任务打造的智能体。 腾讯表示,除了公开基准,它还能在产品系统里,在真实的 AI 研发流水线上,在自然科学研究乃…
Idea2Plan: Exploring AI-Powered Research Planning
AI大模型从想法到研究计划,探索自动化科研规划新范式
arXiv:2510.24891v2 Announce Type: replace-cross Abstract: Large language models (LLMs) have demonstrated significant potential to accelerate scientifi…
Toward Auditable AI Scientists: A Hypothesis Evolution Protocol for LLM Agents
为LLM智能体设计可审计的假设演化协议,让AI科学家过程透明可信
arXiv:2607.09195v1 Announce Type: new Abstract: Large language model (LLM) agents are increasingly expected to play a central role in AI-driven scient…
FirstResearch: Auditable Question Formation for LLM Scientific Discovery Agents
提出可审计的问题形成方法,让LLM科学发现代理更可靠、更透明,研究新思路值得一看
arXiv:2607.05682v1 Announce Type: new Abstract: LLM systems for scientific discovery increasingly assist with ideation, literature synthesis, experime…
Can AI Make Scientific Breakthroughs?
AI能否独立做出科学突破?看顶尖研究者如何辩证这场关于科研范式变迁的精彩对话。
Article URL: https://blog.cosmos-institute.org/p/can-ai-make-scientific-breakthroughs Comments URL: https://news.ycombinator.com/item?id=48775219 Poin…
阿里达摩院:AI 智能体发现 4 种全新超导材料,已获实验验证
AI智能体Elements Claw从6.8万候选物中精准锁定4种全新超导材料,已通过实验验证,改写材料发现效率。
IT之家 7 月 3 日消息,阿里达摩院通过公众号宣布,7 月 3 日(今天),阿里达摩院联合中国人民大学、中国科学院大学等发布首个超导材料发现 AI 智能体 Elements Claw,预测出 6.8 万个可能的超导材料,其中 4 种全新材料已合成并证实存在超导性。相关数据已全部开放。 目前,国际…
LLM-Guided ODE Discovery and Parameter Inference from Small-Cohort Aggregate Data
用大模型引导微分方程发现,从少量聚合数据中反推动力学参数,给科学建模打开新思路。
arXiv:2607.00733v1 Announce Type: cross Abstract: Mechanistic modeling via ordinary differential equations (ODEs) provides interpretable descriptions …
AI systems devise hypotheses and ways to test them
AI不再只是工具,已能独立提出科学假设并设计验证方案,科研范式正在被改写。
Nature, Published online: 30 June 2026; doi:10.1038/d41586-026-01873-2 AI agents collaborate to generate biomedical hypotheses and analyse data, movin…
Ozone loss was a thing even before CFCs were widely used
臭氧层破坏早于人类工业化:这项发现挑战了我们对环境问题的传统认知
With today’s scientific tools, the problem could have been spotted in the 1950s.
Heuresis: Search Strategies for Autonomous AI Research Agents Across Quality, Diversity and Novelty
自主AI研究智能体如何兼顾质量、多样性与新颖性?这篇论文提出全新搜索策略框架,为加速机器学习科研探索提供关键思路。
arXiv:2606.25198v1 Announce Type: new Abstract: Autonomous AI Research promises to accelerate the scientific progress of machine learning. To realise …
特朗普签署行政令:要开发首台“足以开启量子科学发现新时代”的量子计算机
国产量子计算云平台,在线体验真实量子计算机,助力量子算法研究与学习。
IT之家 6 月 24 日消息,美国总统 Donald Trump(唐纳德 · 特朗普)当地时间 22 日签署一项有关量子信息科学与技术 (QIST) 的行政命令,旨在确保美国在该领域保持战略技术优势并构建一个全链路量子生态系统。 ▲ IBM 量子计算 特朗普在行政命令中宣布将 更新国家量子战略 ,…
Discovering Crystal Structure Prediction Algorithms with an AI Co-Scientist
用AI联合作者自主发现晶体结构预测新算法,开启材料科学智能探索新范式
arXiv:2606.22866v1 Announce Type: cross Abstract: We introduce Human-AI Co-discovery system (HACO) for scientific algorithm discovery through cross-do…
A Three-Layer Framework for AI in Scientific Discovery
AI在科学发现中不止搜索与执行,三层框架揭示发现的核心机制。
arXiv:2606.13566v1 Announce Type: new Abstract: Current discussions of AI in scientific discovery are often dominated by two visible capabilities: sea…
Contemporary AI lacks the imagination to diverge or negate in science
最新大规模科学家参与评估揭示,当前AI在科学假设生成中缺乏发散与否定能力,挑战了AI加速科学发现的乐观预言
arXiv:2606.08251v1 Announce Type: cross Abstract: Bold projections that artificial intelligence will accelerate scientific discovery have raced ahead …