Ollama releases6565Ollama now supports Xiaomi MiMo V2.6 (9B parameter code and agent workflow model), as well as AutoTrust decision model (including 9B, 35B, and 27B-VL versions, 35B and 27B-VL can handle images)
MarkTechPostSame story · 4 outlets6262Baidu released ERNIE 5.5 Flash Base, a 280B parameter sparse MoE base model, activating 14B parameters per token, supporting 128K training context and 1M inference context.
Sierra Blog6565Sierra releases AI accelerators, providing preset customer experience base processes (such as order tracking, appointment scheduling, identity verification, etc.) and general integrations, Ghostwriter handles business customization, allowing companies to focus on core business. Accelerators include cross-industry general features (multilingual interaction, satisfaction collection) and industry-specific features (financial domain dialogue guards, medical domain patient authentication), supporting independent updates and cross-system integration, and will open self-built and shared features in the future. The first batch of accelerators has gone live in Agent Studio.
LangChain Deep Agents releases1515LangChain Deep Agents 0.2.5 released, fixes bugs related to matching compressed input budget with selected model, retaining truncated responses when `ModelRefusalMiddleware` throws errors, and applying delays to in-app `uv` installation on Windows.
Codex releases00GitHub Copilot 0.168.0-alpha.7 released, specific update content not announced.
LangChain Deep Agents releases3535LangChain Deep Agents v0.7.33 released, defaulting to error reporting when the model refuses and reporting sub-agent refusal, adding asynchronous sub-agent notification metadata, recording sub-agent checkpoint task call ID. Fixed issues with Python `grep` fallback file range limitation, long result line truncation, and compression failure error reporting, and officially formalized the rubric middleware.
LangChain Deep Agents releases1515LangChain Deep Agents 0.7.34 released, fixed the summary truncation handling issue, and addressed the problem by anchoring the message ID.
Browser UseClaude Code v2.1.159 released, supports 8k context window, performs excellently on SWE-bench Verified; Anthropic launches Browser Use open-source library, allowing AI agents to control browsers via Chrome DevTools Protocol, providing Browser Use Cloud and Browser Use Agents services.
NVIDIA Newsroom5858The NVIDIA Vera processor outperforms the fastest and most energy-efficient x86 CPUs in a 100 megawatt data center comparison, achieving industry-leading agent speed and boosting agent density by 1.6 times per watt. Early adopters like Perplexity and DeepInfra have validated 1.5x and 2.2x agent task acceleration in production testing.
Hugging Face BlogFeatured7171NVIDIA has added four new open-source weight checkpoints for Cosmos 3 under the Cosmos World Model platform, all of which are post-trained versions of Cosmos 3 Nano, containing 16B trainable parameters, using the OpenMDW1.1 license, allowing commercial and non-commercial use.
Why it matters: The original text provides four open-source weight checkpoints and a four-stage training recipe, allowing readers to judge how to adapt the world model to their own sensors and robot bodies.
Hugging Face Daily PapersHalley framework enhances the effective throughput of optical networks in AI data centers through predictive failure and resource control, reducing capacity requirements, and featuring formal guarantees. It includes four components: LightReserve adjusts laser pool scale to provide coverage limits, FoldFit schedules switches within an 11/9 optimal value range, ThermaLead performs one-step advance control on micro-rings, and TriageGuard prioritizes protection for high-risk modules.
TechCrunch · AIOpenAI discovered a severe vulnerability in Ubuntu servers through GPT 5.6 Sol, prompting the company to release the ChatGPT Cyber model for users of the Daybreak Red program, which was announced on November 6, 2026. The model had previously scanned the Linux kernel and found 948 potential issues. OpenAI has also invested $1 billion to support cybersecurity professionals in critical infrastructure, but the industry is concerned about the insufficient allocation of resources for cutting-edge models.
Codex releases00GitHub Copilot 0.168.0-alpha.6 released, specific update content not mentioned.
SemiAnalysisGoogle TPUv7 Ironwood outperforms B200 and B300 by 50% per dollar in single inference performance on Qwen 3.5 397B FP8, and shows competitive or best results in the Multi-Model AgentX benchmark, supporting ecosystems like vLLM and SGLang, but NVIDIA reacts fiercely and accelerates R&D.
The DecoderSame story · 4 outlets
O'Reilly RadarAt the recent Foo Camp, participants believed that the primary carrier of software development is shifting from executable code to natural language prompts, generative processes, and developer skills. Some compared Python to assembly language, suggesting that natural language generation might make traditional code writing optional.
Amp News6565Amp now integrates with GitLab on gitlab.com and self-managed instances, allowing users to start projects from GitLab repositories and enabling orbs to clone and push code。
Databricks Blog5858Databricks and the Linux Foundation announced the upgrade of Delta Sharing to OpenSharing in June 2026, expanding the sharing scope from Delta Lake tables to include Agent skills.
Mastra BlogThe Mastra platform and Studio now support tracking latency metrics for each agent and model call, with filtering and sorting of tracking, span, and threads via TTFC (First Data Block Time) and TTFT (First Token Time) to identify slow components.
OpenClaw BlogOpenClaw Multiplayer allows multiple people to share the same agent and context to collaborate without reconstructing the dialogue or passing instructions, supporting over 50 people collaborating in the same environment.
Codex releases00GitHub Copilot rust-v0.168.0-alpha.5 has been released, this version provides code review, collaboration, and other functions.
Microsoft ResearchThe PDD Kit translates user intent into formal specifications and verifies whether the code written by the agent for each input conforms to these specifications, supplemented by testing. The team officially validated the Rust-written Pet Store login logic using the Chopin environment on GitHub Next with the PDD Kit, and the Verus tool discovered and fixed existing bugs, proving no login bugs under the specification.
Mastra Blog4242Mastra launches a customization kit program where engineers will embed with customer teams to build kits that meet their goals; Workday migration to Mastra completed in just 31 days, reducing 3,000 lines of code and lowering vector storage costs by 98%; customers can avoid vendor lock-in and achieve measurable ROI.
NVIDIA NewsroomChris Alexiuk promotes the open-source multimodal model NVIDIA Nemotron, supporting various versions from Nemotron 3.5 Lightning to Nemotron Ultra, helping developers build LLM applications and agents. He makes AI easy to understand and use through YouTube tutorials, emphasizing the importance of full open-source for democratizing AI.
NVIDIA Technical Blog6565NVIDIA Co-Design optimizes memory usage by reducing redundant weight replication, idle KV cache, and repeated prefix computation, enhancing HBM efficiency and allocating more memory to token generation. This approach enables each bit to perform more useful work rather than solely focusing on GPU memory capacity.
LangChain BlogBrowserbase Stagehand tool enables agents to complete multi-step browser operations with a single call, combining snapshot view management to context, significantly improving Claude Sonnet 5.5's pass rate in the WebVoyager 100 task by 90% (total cost $1.98).
LangChain BlogLangChain explores three principles of human feedback to enhance agent development: expert judgment defines quality standards, comprehensive context evaluation, and converting feedback into reusable data; through agent trace to record user interactions and model behavior, manually labeled scoring with reasons, used to calibrate automated evaluators, improve evaluation criteria, and build golden datasets, forming reusable expert knowledge assets.
LangChain Blog5252LangSmith Bring Your Own Cloud (BYOC) now supports AWS (GA), Azure (Beta), and Google Cloud (Beta), allowing users to keep tracking, evaluation, datasets, prompts, and agent deployments in their chosen cloud accounts, with LangChain managing the data layer.
Hugging Face Daily Papers6565EmoReward adapts multi-reward weight allocation by extracting evidence reliability from reward trajectories collected during actual training runs to calibrate reward weights for emotional video generation. This mechanism first evaluates specific grounding cues based on emotional prompts, then scores videos during the generation phase using heterogeneous signal primitives, and calibrates their contributions through reflective reward gating.
Ripple-65B Improves Computational Efficiency by Optimizing SFT and RL Schemes, Using Synthetic Data Pipelines and Quality Filtering, Employing a Hybrid RL Scheme with Truncated Importance Sampling, Score Centering, and Linear Length Penalty in a High-Quality Long-Term Synthetic Environment, Achieving a TerminalBench 2.1 Score of 80.5 with Only 1.4M Rollouts, 70 Times Fewer than Reflection Beam
ByteDance Seed · Hugging Face new modelsByteDance-Seed/Qwen3-30B-A3B-MoE-SAE is released, containing TopK sparse autoencoders (k=100, 2^20 latent variables) trained on layers 4, 30, and 46 of the Qwen3-30B-A3B model. Each SAE is approximately 17.2 GB, with reconstruction variance unexplained scores of 12.6%, 25.1%, and 7.6% respectively.
Databricks BlogNVIDIA and Databricks jointly contributed the Zerobus Ingest message receiver for the Debezium project, which writes records directly to Unity Catalog Delta tables.
IBM Research Blog3636ORNL, IBM, and AMD deploy a quantum error correction (QEC) collaborative design platform at IBM's New York Poughkeepsie campus, integrating the IBM Quantum Nighthawk r2 processor with the ORNL Umbra HPC system, which includes the AMD Instinct MI355X GPU and 5th Gen AMD EPYC™ CPU, providing low-latency quantum-classical execution capabilities for algorithm development and validation in hybrid HPC-quantum computing environments.
IBM Research Blog3838HRL Laboratory develops an inertial measurement unit (IMU) using advanced algorithms, providing an uninterrupted GPS alternative. Its MEMS inertial sensor is the size of a can of soup, with accuracy close to GPS, and solves the long-standing problem of bias drift. Through the AutoCal automatic calibration system, it continuously corrects errors, achieving unlimited navigation support. The IMU uses ASIC and advanced packaging technology to optimize cost and size, making it suitable for light vehicles such as drones.
Hugging Face Daily PapersStreamDuo pairs a trainable commentator with a frozen research agent that searches for game wiki entries to improve cross-game live commentary without training specifically for the target game. StreamDuo was trained on Baldur's Gate 3 and tested on ten unseen new games, increasing entity recall from 0.082 to 0.116, reducing source game leakage from 9.9% to 1.0%, and achieving the best timing, highest evaluation vitality, and overall quality in the live model.
O'Reilly RadarRouteLLM researchers train routers to choose between strong and cheap models, benchmark tests show more than double cost reduction with minimal impact on response quality. Automatic routing can help users get more help within budget, allowing the same subscription to support more requests or more frequent use of strong models. Users need to understand the routing effect and evaluate the cost and time benefits it brings, such as comparing different routing strategies through full interaction of code analysis tasks.
Codex releases00GitHub Copilot rust-v0.168.0-alpha.4 released, this version includes the latest feature improvements and fixes.
Hugging Face BlogFeatured7272Phonon releases a small TTS model that can run on local devices, without a server and network, offering two versions of about 65 MB and 75 MB, covering English, French, German, Spanish, and Portuguese.
Why it matters: The original text outlines the specifications, open-source license, and leaderboard performance of on-device TTS models, allowing judgment on the usability boundary of small models in local voice scenarios.
