文章 RSS Feed

• 2026/04/17

  使用 Sentence Transformers 训练和微调多模态嵌入与重排序模型

  https://huggingface.co/blog/train-multimodal-sentence-transformers
• 2026/04/16

  Gemini Robotics-ER 1.6：通过增强的具身推理驱动真实世界机器人任务

  https://deepmind.google/blog/gemini-robotics-er-1-6/
• 2026/04/16

  在 AWS Trainium 和 vLLM 上使用推测解码加速解码密集型 LLM 推理

  https://aws.amazon.com/blogs/machine-learning/accelerating-decode-heavy-llm-inference-with-speculative-decoding-on-aws-trainium-and-vllm/
• 2026/04/16

  深入 VAKRA：Agent 的推理、工具使用与失败模式分析

  https://huggingface.co/blog/ibm-research/vakra-benchmark-analysis
• 2026/04/15

  Amazon Bedrock AgentCore 的 Spring AI SDK 正式发布（GA）

  https://aws.amazon.com/blogs/machine-learning/spring-ai-sdk-for-amazon-bedrock-agentcore-is-now-generally-available/
• 2026/04/15

  自动化弱到强研究者（Automated Weak-to-Strong Researcher）

  https://alignment.anthropic.com/2026/automated-w2s-researcher/
• 2026/04/15

  在 Amazon SageMaker HyperPod 上运行推理的最佳实践

  https://aws.amazon.com/blogs/machine-learning/best-practices-to-run-inference-on-amazon-sagemaker-hyperpod/
• 2026/04/14

  如何使用 AWS Lambda 为 Amazon Nova 模型定制构建有效的奖励函数

  https://aws.amazon.com/blogs/machine-learning/how-to-build-effective-reward-functions-with-aws-lambda-for-amazon-nova-model-customization/
• 2026/04/14

  我们如何破解顶级 AI Agent 基准测试：以及下一步该怎么做

  https://rdi.berkeley.edu/blog/trustworthy-benchmarks-cont/
• 2026/04/13

  在 App Service 上使用 OpenTelemetry 和 Application Insights 新 Agents 视图监控 AI Agent

  https://techcommunity.microsoft.com/blog/appsonazureblog/monitor-ai-agents-on-app-service-with-opentelemetry-and-the-new-application-insig/4510104
• 2026/04/13

  在 Azure App Service 上使用 Microsoft Agent Framework 1.0 构建多 Agent AI 应用

  https://techcommunity.microsoft.com/blog/appsonazureblog/build-multi-agent-ai-apps-on-azure-app-service-with-microsoft-agent-framework-1-/4510017
• 2026/04/11

  Agent 治理工具包：架构深度探讨、策略引擎、信任和 AI Agent 的 SRE

  https://techcommunity.microsoft.com/blog/linuxandopensourceblog/agent-governance-toolkit-architecture-deep-dive-policy-engines-trust-and-sre-for/4510105
• 2026/04/11

  介绍 Muse Spark：迈向个人超级智能的扩展

  https://ai.meta.com/blog/introducing-muse-spark-msl/
• 2026/04/11

  实践中的可信赖 Agent

  https://www.anthropic.com/research/trustworthy-agents
• 2026/04/10

  更好的 Harness：使用 Evals 进行 Harness Hill-Climbing 的秘诀

  https://blog.langchain.com/better-harness-a-recipe-for-harness-hill-climbing-with-evals/
• 2026/04/10

  Agent 改进循环中的人类判断

  https://blog.langchain.com/human-judgment-in-the-agent-improvement-loop/
• 2026/04/10

  使用 Sentence Transformers 的多模态嵌入与重排序模型

  https://huggingface.co/blog/multimodal-sentence-transformers
• 2026/04/09

  ALTK-Evolve：AI 代理的在岗学习

  https://huggingface.co/blog/ibm-research/altk-evolve
• 2026/04/09

  追踪每一个 Token：Microsoft Foundry Agents 的粒度成本和使用指标

  https://techcommunity.microsoft.com/blog/azure-ai-foundry-blog/tracking-every-token-granular-cost-and-usage-metrics-for-microsoft-foundry-agent/4503143
• 2026/04/08

  Claude Mythos 预览版：网络安全能力评估

  https://red.anthropic.com/2026/mythos-preview/