Data & Training News

Latest reporting, research and product updates filed under Data & Training.

25 picksNewest firstLatest article Sep 28, 2026, 5:00 PM

Latest stories

October 1, 20261 story

Modal Clusters are generally available

Modal 宣布 Modal Clusters 正式可用，通过一个装饰器 @modal.clustered 即可获得多节点集群，节点间经 InfiniBand verbs 通信可达 6.4 Tbps，自动配置 PyTorch 和 NCCL。

September 28, 20261 story

17:00MIT News（RSS）AI score 78/100

New formulation helps RNA vaccines withstand high temperatures

RNA vaccines, which have been proven effective against Covid-19, are now being developed for many other diseases, including cancer. One of the drawbacks to these vaccines is that they require ultracold storage, but researchers from MIT have found a promising way to overcome that limitation.With help from an AI algorithm, the researchers tweaked the formulation surrounding the lipid nanoparticles that are typically used to deliver mRNA vaccines, making the vaccines more heat-resistant. Using this

September 27, 20261 story

00:00小米 MiMo：官网发布与博客AI score 63/100

Diagnosing and Mitigating Tool-Call Repetition in MiMo-V2.6

September 27, 2026A lesson from scaling RL: the reward blind spot in optimizing for correctnessFollowing the release of MiMo-V2.6, tool-call repetition emerged as one of the most noticeable issues affecting the user experience. In MiMo Desktop, MiMo Code, OpenCode, and other agentic settings, the model would sometimes issue the same or highly similar tool calls repeatedly, consuming substantial time and context without making meaningful progress.Our internal evaluations confirmed this pattern: t

September 23, 20261 story

00:00Anthropic：Newsroom（网页）AI score 76/100

Claude discovers a novel enzyme system with CRISPR-like repeats

Anthropic 宣布成立生命科学研究组与实验室，Claude 智能体在仅有人类高层指导下自主发现一种与 DNA 重复序列相关、结构类似 CRISPR 的新型酶系统，命名为 array-associated reverse transcriptases（ART），并发布预印本。

September 12, 20261 story

00:28Dwarkesh Patel：Podcast & Blog（RSS）AI score 67/100

AI researchers debate how close we are to recursive self-improvement

Dwarkesh Patel 与 Zyphra CTO Beren Millidge、Thinking Machines 首席科学家 John Schulman、Baseten 模型训练负责人 Charlie O'Neill 三位研究者对谈递归自我改进（RSI）的前景。

September 11, 20261 story

08:00Together AI 研究与产品博客（RSS）AI score 63/100

Together AI expands fine-tuning service with more models, live metrics, and finer controls

Together AI 扩展 Together Fine-Tuning 服务，新增对 GLM-5.3、Kimi K2.7、Qwen 3.5 系列等最新开源权重模型的支持，并上线实时实验跟踪、数据集预览验证等能力。

September 10, 20261 story

00:00Cognition 模型 / Devin 博客（网页）AI score 70/100

Introducing SWE-2: Pushing the Pareto Frontier

Cognition 发布最先进编码模型 SWE-2，基于 2.8T 参数的 Kimi K33 后训练，在 FrontierCode 1.1 Main 取得 50.0%，仅比 Fable 5.1 低不到一分且便宜 64%。

September 9, 20261 story

21:27Anthropic：The Institute（旗舰研究长文 · 网页）AI score 64/100

Scenarios for our Economic Future

Anthropic 经济团队基于技术报告《Economic Scenarios for Transformative AI》发布交互式经济情景探索工具，把经济表示为任务束，推演 AI 到 2030 年对美国经济的影响。

September 3, 20262 stories

08:00Tomer Tunguz 博客（VC 分析）AI score 73/100

The Ads Model for Prompts Vertically Integrates AI

Tom Tunguz 分析 Meta 发布 Muse Spark 模型及双轨 API 定价：Standard Tier（muse-spark-1.3）输入 $1.25/m。

08:00Hugging Face：Blog（RSS）AI score 63/100

Training a coding model to paint watercolours with TRL and OpenEnv

Hugging Face 博客作者基于 Surya Narreddi 的原始想法，用 TRL、OpenEnv 和 Qwen/Qwen3.5-35B-A3B 复现了让语言模型通过 p5.brush 写 JavaScript 画水彩的 RL 训练流程，所有数据集、环境、脚本和模型均开源在 Hub。

August 26, 20261 story

08:00Hugging Face：Blog（RSS）AI score 73/100

Training and Finetuning Multi-Vector Embedding Models with Sentence Transformers

Sentence Transformers v6.0 新增第四种模型类型 MultiVectorEncoder，支持 ColBERT 风格的后交互检索，并配套完整训练流程。

August 13, 20261 story

08:00Epoch AI：研究、数据与评测AI score 61/100

The performance per dollar of AI chips purchased each quarter has grown by an average of 49% per year

Epoch AI 估算 2023 年 Q1 至 2025 年 Q4 每季度购买的 AI 芯片综合性价比，从约 5.6×10^11 升至约 1.4×10^12 bit-operations per second per dollar（2025 年不变美元），指数拟合平均每年增长约 49%（90% CI：36%–66%），倍增时间 1.7 年。

Showing the latest 12 of 25 stories.