Tag
This arXiv paper shows how 'warped formatters' — tool output text that omits key details like search limits or result caps — mislead AI agents and their verifiers even when the underlying computation is correct. Repaired formatters drastically cut unsupported claims and flipped verifier acceptance from 90/90 wrong to 90/90 correct, and an audit prompt rediscovered live defects in the authors' own system plus OpenClaw and Hermes, prompting merged fixes.
SPECMEM提出了一种在部署期间持续维护的工具调用记忆系统，通过在线写回已验证的调用记录并配合LRU淘汰和副作用门控机制，在多会话多用户场景下实现最高1.74倍的墙钟加速，同时避免了不可逆操作的错误。
论文提出 The Trinity 框架，通过 Creator、Explorer、Evolver 三个角色将多模态智能体的递归自我改进组织成能力树与 Harness 树，在 28 个训练基准上平均提升 34 分，18 个未见基准上提升 26 分。
Anthropic 将 Claude Cowork 与聊天功能统一后的体验正式向 Claude Enterprise 客户开放，支持跨网页、桌面和移动端的云端代理工作，并新增 Efficient mode 等管理员成本控制功能。
Nanbeige4.3-33B 是南ベージ系列的旗舰教师模型，宣称在约30B参数级别的开源模型中达到最先进水平，专注于复杂编码与专业协作智能体任务，并为后续小模型蒸馏（如 Nanbeige4.3-3B）奠定基础。
Liquid AI released LFM3-1.7B, a 1.73B-parameter tri-hybrid text model designed for local agents and edge deployment with a 256K-token context window, fitting in about 1 GB of memory at 4-bit precision and supporting single-GPU fine-tuning for domain specialization.
Apple 的 Mac Mini M4 基础款降价 150 美元至 649 美元，但考虑到其 2024 年首发价仅 599 美元（后因芯片与内存短缺涨至 799 美元），这笔折扣的实际价值存疑。文章指出，对不需要 M6 更强性能的大多数用户而言，这仍是不错的选择。
Box CEO Aaron Levie argues that cybersecurity will be a major source of AI-generated jobs and one of the biggest enterprise investment areas, even as work becomes more agentic and autonomous, citing critical PHP vulnerabilities as a timely example.
PyTorch Con 2026 主题演讲预告：Agentic AI Foundation 执行董事 Mazin Gilbert 探讨 LLM 如何从智能体行为中学习，提出五层智能体栈架构，并将智能体评估反馈为 PyTorch 训练奖励的开放研究方向。
A research post introducing 'reactive optimizers': an agent watches training dynamics and rewrites the optimizer's source code during training, matching or beating the best hand-tuned recipes on image and language tasks while autonomously guarding against loss spikes and overfitting. The method is released via the open-source OPAL package.
CrowdStrike researchers found four previously unreported threat groups using agentic AI, primarily older Claude and DeepSeek models, to run autonomous hacking campaigns, circumventing model guardrails through jailbreaking and fake authorization files to exploit basic security hygiene gaps at scale.
NVIDIA announced an expanded family of open models across language, robotics, biology and quantum computing—including Nemotron 3.5 Super VL/Bolt, Nemotron Retriever, Kumo, Cosmos 3.5, Alpamayo 2 Edge, BioNeMo and Ising—to help European organizations build sovereign, locally customized AI applications.
在 GTC Berlin 2026 主题演讲中，NVIDIA CEO 黄仁勋宣布 DSX POD 与 DSX AI 工厂平台等新品，强调欧洲在 AI 经济中的角色，并将智能体（agents）视为新的计算平台。
Menlo Research announces Asimov 1, an open-source humanoid robot built from off-the-shelf parts for agentic robotics, with open CAD, SDK, safety layer, and locomotion policy so agents can understand, repair, and extend the robot.
InternScience 发布 Sage-A1.5，一个基于 Qwen3.6-35B-A3B 微调的 35B MoE 智能体模型，主打长程科学问题求解，具备智能体编排、自我验证与开放式科学发现能力，在多项基准上与 Kimi-K3、GLM-5.3、GPT-5.6-sol 等前沿模型竞争。
本文系统综述了由大语言模型驱动的工业智能体，分析关键挑战，提出统一分类法、四层系统架构，并回顾其在智能制造全生命周期中的应用与未来研究方向。
The paper presents MedPAO-Assist, an edge-deployable, modality-aware agentic AI framework that structures radiology reports from dictation using a ~0.6B-parameter small language model, running with ~3.3GB RAM and ~31s latency on a CPU-only laptop.
The paper introduces AW-0, an agentic world action model framework that unifies code-based planning with joint video and action prediction in a mixture-of-transformers architecture, enabling reusable skill libraries for robot manipulation. Evaluations on RoboTwin2.0 and bimanual manipulation tasks show improved performance over VLA and WAM baselines.
Ujval Kapasi of NVIDIA delivers a PyTorchCon 2026 keynote on what agentic workloads demand from the PyTorch stack, using NVIDIA Vera Rubin as a case study for hardware-software co-design across CPUs, GPUs, memory, and networking, covering optimizations in PyTorch, Triton, and vLLM.
Tim O'Reilly hosts a live conversation with Alex Komoroske about his Resonant Computing Manifesto and his new project Common Fabric, a human-centered approach to the agentic future.