Claude Opus 5.5 fast mode 结束研究预览并移除等待名单
AIHOT ·
Anthropic 宣布 Claude 平台上的 Opus 5.5 fast mode 不再处于研究预览阶段，并已移除等待名单。该模式速度最高提升 2.5 倍，定价为 $8 / $40 per million tokens；现已登陆 Microsoft Foundry（Azure 托管），Google Cloud 即将支持。Claude Code 中的…
原文内容 · 暂无该语言译文AI News · 2026-10-15
按本站收录日期（UTC）归档，不代表事件发生于当天；原公告、聚合源发布与收录时间分别标注。
返回资讯列表Meta 开放权重模型 Muse Spark Open 上架 OpenRouter
AIHOT ·
Meta 发布开放权重模型 Muse Spark Open，现已可在 OpenRouter 使用，定位为面向编码与智能体工作的美国最强开放权重模型。模型总参数 816B、激活 42B，支持 1M token 上下文窗口，原生多模态，OpenRouter 定价为输入 $1.25/M、输出 $4.25/M，正文提到支持图像、视频、音频和 PDF 输入。
原文内容 · 暂无该语言译文Safe Reinforcement Learning with Receding-Horizon Safety Critic and Differentiable Action Shields
arXiv cs.RO ·
Despite the Markov assumption, safety failures in robot reinforcement learning rarely occur instantaneously; they typically emerge over several steps as the policy enters unfami…
原文内容 · 暂无该语言译文LintEval: Benchmarking LLMs on Linter-Guided File Refactoring
arXiv cs.CL ·
Linters are widely used in code maintenance and real-time IDE editing to surface lint alerts related to performance, style, error-prone, and other patterns in source files. Howe…
原文内容 · 暂无该语言译文TrAct: Bridging Robot Control and Visual Prediction with Visual Tracks
arXiv cs.RO ·
Robot actions are inherently embodiment-specific and only weakly aligned with image-space visual changes, limiting their effectiveness as conditioning signals for robot world mo…
原文内容 · 暂无该语言译文PersEvalBench: Can LLM Evaluators Judge Contextually Appropriate Use of Persona Information?
arXiv cs.CL ·
Large language model (LLM) evaluators are increasingly used to assess personalized assistants, yet whether they can reliably judge personalization has rarely been tested. We int…
原文内容 · 暂无该语言译文Claude Code 空闲压缩从摘要继续，可用 idleCompaction 关闭并自定义阈值
AIHOT ·
Claude Code 的空闲压缩会让 Claude 从摘要而非完整上下文继续。作者说明可在 settings. 中设置 "idleCompaction": false 关闭，并通过 CLAUDE_CODE_IDLE_COMPACT_MIN_TOKENS 自定义阈值，默认为 200k。
原文内容 · 暂无该语言译文Tree Appreciator
Simon Willison ·
Tool: Tree Appreciator Another game experiment with Claude. This one used the default Claude iPhone app (chat mode, not Claude Code mode), where I prompted Claude Opus 5.5 Max w…
原文内容 · 暂无该语言译文Databricks 讲解如何在 AI Runtime 上将 Qwen 3.8 27B 微调时间缩短 6 倍
AIHOT ·
Databricks 演示通过系统性性能优化，将 Qwen 3.8 27B 在 AI Runtime 上的微调时间从 7 小时 44 分钟降至 1 小时 17 分钟，提速 6 倍。
原文内容 · 暂无该语言译文Google DeepMind and Singapore Government collaborate to build fine-tuned AI model for scam detection
Google DeepMind ·
Learn how Google DeepMind and GovTech Singapore uses lightweight on-device Gemma models to intercept conversational scams in private chats with zero cloud data egress.
原文内容 · 暂无该语言译文