vLLM BlogFeatured6060The vLLM-Omni team has integrated MiniMax H3-Image into NVIDIA Vera Rubin. With a four-card SP4 configuration, the native 4096×4096 image-to-image short prompt benchmark median latency is 0.87 seconds, and the throughput of four copies SP1×DP4 reaches 6115 images per hour.
AWS Machine Learning BlogJava teams can leverage Amazon Bedrock and Spring AI to build autonomous AI agents, addressing scalability, governance, security, and operational challenges. Spring AI brings AI capabilities to the JVM through the Spring framework (auto-configuration, dependency injection, composable advisors), while the Amazon Bedrock AgentCore platform supports framework and model-agnostic agent construction, connection, and optimization.
AWS Machine Learning BlogDevelopers use OpenAI Codex for code exploration and engineering work through Amazon Bedrock. Amazon Bedrock supports Codex inference and can be integrated with AWS Identity, IAM, and observability practices. It is recommended to start with Native AWS Access for small-scale pilots, use GPT-5.6 Sol for validation, authenticate through IAM Identity Center, and configure model permissions and observability targets.
Apple Machine Learning ResearchWeighted State Souping proposes query dependency weighting to combine cached State state encodings, significantly improving long-context performance on tasks such as single-hop and multi-hop document question answering, pinpoint retrieval, variable tracking, and tool use. The method can be adapted into an efficient sparse variant, achieving near-constant encoding costs while maintaining (or sometimes improving) performance.
Apple Machine Learning ResearchState Space Models (SSMs) lag behind Transformers in contextual learning and precise retrieval tasks, as their gating mechanism leads them to first learn memory solutions, delaying or preventing the convergence of correct contextual learning solutions, but the gating mechanism is beneficial for improving long sequence generalization.
Apple Machine Learning ResearchResearchers evaluated the performance of Large Language Model (LLM) judges in personalized writing tasks and found that the quality of judges is weakly correlated with general LLM capabilities measured by MMLU (Pearson’s ρ ≤ 0.2), and their performance is influenced by writing tasks and evaluation criteria (e.g., general helpfulness vs. stylistic personalization). LLM evaluators were more consistent and performed better in pairwise comparisons than in independent scoring responses. The 'strongest' LLM did not reliably replicate human evaluations, possibly due to being overly sensitive to subtle details overlooked by human annotators and weaker LLMs.
Claude · YouTubeSame story · 3 outlets4242Claude Code released Fable 5.2, requiring users to select Opus 5.5, Fable 5.2, Sonnet 5.5, or Haiku 5.5 based on task type, or distribute tasks across models. The availability and usage restrictions of each model vary depending on the subscription plan. Guidelines suggest using the corresponding model for complex coding and knowledge work, open-ended technical tasks, and low-cost limited tasks, respectively.
How the accounts differ:TechCrunch states the input price of Fable is reduced to $8 per million tokens; others do not mention the price.
Amazon ScienceAmazon Web Services uses its automated reasoning tool Tiros to verify network isolation by converting network configurations into logical propositions and analyzing possible paths in the model to check if isolation complies with stated policies, replacing traditional dynamic sampling methods.
Meta Engineering4040Meta's Intelligent Delivery system, through AI optimization of computational resource allocation, has boosted ad sorting performance by approximately 10% since 2017. The system includes a Resource Core Model that predicts the marginal utility of computational resources (6B sparse parameters + 20M dense parameters), multi-layer mechanisms for cross-system resource migration, and a framework that uses AI agents to continuously monitor and detect changes in the response system.
AWS Machine Learning BlogAWS Quick's AI-driven product development lifecycle (AI-PDLC) transforms enterprise signals into validated prototypes through six specialized skill agents, combined with human approval. The Discovery agent analyzes pain points in enterprise data sources, and the Prioritize agent uses a 0-10 score framework to evaluate use case priorities, with all decisions requiring manual confirmation.
AWS Machine Learning BlogAmazon SageMaker AI Two-Way Stream introduces a pay-as-you-go pricing model, enabling model providers to report actual usage of real-time streaming inference through the AWS Marketplace. Deepgram becomes the first partner to bring voice-to-text and text-to-speech pay-as-you-go to the AWS Marketplace.
Databricks Blog5858Databricks releases Genie Ontology, an enterprise context layer, combining managed definitions with continuous learning of organizational knowledge to provide business understanding capabilities for AI agents. Genie plugins support ChatGPT, Claude, Microsoft Copilot Cowork, and Grok Bot, enabling teams to maintain business context in Databricks and use it across multiple agents.
Databricks Blog5858Databricks Genie Agents can now perform high-quality business analysis, analyze large volumes of documents, and connect enterprise tools to automate actions by supporting unstructured data analysis and new MCP connectors. The new OpenSharing feature allows users to share managed agents with pre-installed data, instructions, and knowledge with partners, suppliers, and customers, accelerating the generation of external insights.
Hugging Face Blog3838Hugging Face's blog introduces the Token Pooling perception training method for multi-vector embedding models, which compresses storage by merging similar tokens through HierarchicalTokenPooling.
vLLM Blog5858vLLM extends multi-modal service capabilities through the EPD (Encoder/Prefill/Decode) separation technology, supporting independently deployed visual encoders. vLLM v0.32.0 includes MiMo and XPU support, and the metadata-only request feature of DeepSeek V4.1 Flash has been open-sourced.
Codex releases00GitHub Copilot 0.169.0-alpha.11 has been released, with no specific updates or performance metrics mentioned in this version.
Ant Group inclusionAI · Hugging Face new modelsinclusionAI has released the ZoomGain-4B large language model, achieving a score of 61.08 on ZoomGainBench and performing excellently on multiple fine-grained perception and general perception benchmark tests; the model is built with supervision based on the global-to-local visual perception gap mined from global image details, supporting English and Chinese Q&A, and can be downloaded via Hugging Face.
Ant Group inclusionAI · Hugging Face new modelsSing VL S1 is a multimodal decision-making model based on Qwen3.5, capable of answering named entity questions with text or JSON and up to four images, scoring options for each question using a single Transformers forward pass, performing excellently on 43 text datasets and 11 visual datasets, supporting Python 3.12 and PyTorch inference.
Hugging Face BlogTether AI Research has launched the 0.8B parameter BotPsy model, built based on Qwen3.5-0.8B, specifically designed for function calls and agent workflows, suitable for on-device deployment. BotPsy outperforms all ≤1.2B parameter models in 6 tool use and agent benchmark tests, including Qwen3.5-2B-Think, achieving the highest scores in 5 tests, with a 2.5x speed increase and a 3x improvement in reasoning efficiency.
NVIDIA BlogThe NVIDIA Jetson Orin NX module is used for inference on the pilot line of AIRECO's recycling plant, driving a robotic arm to pick up target objects from a conveyor belt with an accuracy rate of 95%; plans are to increase the annual processing capacity from 18,000 tons to 48,000 tons next year, reducing operating costs by 85%.
AWS Machine Learning BlogMoonshot AI's Kimi K3.1 Pro Highspeed has been released on Amazon Bedrock, the model being a post-training version of the Kimi K3.
Microsoft ResearchMicrosoft researchers discover 'Lazarus' technique can reprogram AI model dormant neurons to plant backdoors, enabling normal AI agents to execute malicious commands under trigger conditions. The technique can plant backdoors without retraining and is difficult to detect through conventional tests, posing a particularly severe threat to AI agents with tool or terminal access.
AWS Machine Learning BlogDevelopers used LangGraph and Amazon Bedrock AgentCore Web Search to build a daily dependency monitor, watching daily changes in open-source software packages and container base images. The monitor checks daily new impacts on team releases, searches using Anthropic Claude models, and uses the Web Search API to get current results.
PyTorch BlogPyTorch models run on WebGPU in browsers via ExecuTorch without converting framework formats. Export as ExecuTorch .pte files and run on any WebGPU browser using the new ExecuTorch WebGPU backend.
AWS Machine Learning Blog4040Amazon Bedrock AgentCore Evaluations provides three evaluators—built-in, DeepEval, and AutoEval—to assess agents. The built-in evaluator is developed and tested by AWS, covering common dimensions; DeepEval and AutoEval are third-party open-source libraries, with AgentCore hosting them in production environments. The evaluation process requires first defining business success criteria, breaking them down into observable behaviors, and then mapping them to the appropriate evaluator level.
OpenAI Alignment BlogGPT-6 Astra shows a decrease in CoT monitoring ability in non-adversarial settings, an increase in CoT evasion ability in adversarial tests, and no significant improvement in CoT capability. During pre-training, CoT capability does not increase with the decrease of the base model's perplexity, and the architecture has a limited impact. Reinforcement learning enhances CoT control ability, and model-generated data fine-tuning can transfer CoT control ability. Model perplexity is positively correlated with CoT capability and control ability.
LangChain Deep Agents releases3535LangChain Deep Agents deepagents-code v0.2.7 released, adding shell completion scripts, fixes Windows editor launcher detection, `dcode completion` output, and Bash completion issues, improves Git attribution performance and adds a disable option.
Google AI BlogGemini Intelligence is now available on more devices, including the Motorola Signature 27 and Pixel 10. New features will expand capabilities on these devices, helping users complete more tasks. Create My Widget is launching soon, allowing users to create personalized widgets with prompts; task automation will expand to the UK, India, and Japan.
AWS Machine Learning BlogCodex releases00GitHub Copilot 0.167.1 is released, providing features such as code review, vulnerability detection, and secret scanning.
Google Developers BlogGoogle AI Edge Gallery has launched on-device automatic speech recognition (ASR) and text-to-speech (TTS) capabilities, enabling fully offline operation of multilingual ASR and natural TTS models on phones and wearable devices.
Meta EngineeringMeta open-sources Triton Low-Level Extensions (TLX) compiler, enhancing GPU performance by introducing explicit warp-group and cluster-level orchestration, supports RecSys training/inference and LLM pre-training, achieving end-to-end performance improvements over standalone libraries on NVIDIA and AMD GPUs.
Meta EngineeringMeta has launched the Your Algorithm tool, allowing users to edit interest lists ("Interested In"/"Not Interested In") in natural language to adjust Feed and Reels recommendations in real-time. The system uses LLM to process user preferences and integrate user behavior, achieving responsive adjustments through retrieval, reordering, and pushing optimization stages. New interest content will appear quickly but will not dominate recommendations.
LangChain BlogWorkday has utilized LangGraph for agent orchestration and LangSmith for tracking and evaluation, reducing prototype development time from 5 months to within 1 day with a 4-person team by 2025 and supporting production environment deployment. Its Python SDK 'Workday LangChain' encapsulates LangChain core components and routes through internal services (including LLM gateway, RAG retrieval) while meeting enterprise-level security and compliance requirements.
PyTorch BlogExecuTorch WebGPU backend allows PyTorch models to run in WebGPU browsers without converting framework formats, supporting multiple models such as Llama 3.2, Qwen3, and Gemma 3.
AWS Machine Learning BlogTCS has built an Agentic test orchestrator for BFSI clients, transforming requirements into executable controlled tests through Amazon Bedrock AgentCore, breaking down tests into dedicated tasks and executing them using their respective agents, tools, and validation rules, while meeting project-specific context, API, and browser UI testing, tenant-level security isolation, and repeatable test path requirements, significantly improving test throughput and allowing engineers to focus on risk analysis and exploratory testing.
Codex releases00GitHub Copilot released version 0.169.0-alpha.9, without mentioning specific update content.
