周报

周报

2026-W33 周报精选：The Defender’s Window；AiToEarn：Let's use AI to Ear；OpenAI joins PORTS-Pike proj

条目
80
覆盖天数
7
当前批次
2026-W33
往期
#07

豆包更新了工作模式，正面硬刚 Workbuddy

AI 办公场景现在成了兵家必争之地。 字节终于出手，豆包客户端最近更新，增加了全新的工作任务模式，体验跟 Codex 非常接近。 能用浏览器、能调 Skill，内置 Office 和飞书办公套件，同时支持本地和云端电脑，合上笔记本也能继续工作。 测试了 7-8 个任务，竟然有点超预期，宇宙厂终于在办公 Agent 场景

推荐理由「豆包更新了工作模式，正面硬刚 Workbuddy」可当助手或自动化原型来拆，看交互和工具边界，不是发布会。
创作与自媒体乔木博客
blog.qiaomu.ai打开原文
#57

Can We Defend Against AI-Generated Video Attacks on Real-World Crisis Events? A Systematic Evaluation of Detectors, Generators and Social Di

Recent video generators can fabricate realistic depictions of wars, disasters, public emergencies, and other real-world crises, creating substantial risks of misinformation. Existing benchmarks, however, provide limited

推荐理由「Can We Defend Against AI-Generat…」对造物者有用的是方法能不能迁到自己的 Agent 或生成管线，不是把标题再念一遍。
研究与学习HF daily papers
huggingface.co打开原文
#59

Apodex Discovery: Reality Benchmarks and Environments for Evaluating and Building Discoverative Artificial Intelligence

Apollo did not reach the Moon merely because its engineers could solve difficult equations. It succeeded by turning a distant ambition into a mission architecture of explicit objectives, simulation, verification, and rep

推荐理由「Apodex Discovery: Reality Benchm…」对造物者有用的是方法能不能迁到自己的 Agent 或生成管线，不是把标题再念一遍。
研究与学习HF daily papers
huggingface.co打开原文
#60

Beyond Final Scores: A Systematic Evaluation of Agents for Long-Horizon AI Research and Development

Autonomous agents are increasingly capable of improving models, systems, and other technical artifacts through long-horizon experimentation. To understand the current state of this capability, however, evaluation must go

推荐理由「Beyond Final Scores: A Systemati…」偏Agent，对造物者有用的是方法能不能迁到自己的 Agent 或生成管线，不是把标题再念一遍。
研究与学习HF daily papers
huggingface.co打开原文
#61

SimpleOPD: Simple Tokenizer-Agnostic On-Policy Distillation for Long-Context Reasoning

On-policy distillation (OPD) offers a promising way to transfer reasoning capabilities from stronger teacher models, but applying it to long-context reasoning teachers and short-context students introduces practical chal

推荐理由「SimpleOPD: Simple Tokenizer-Agno…」对造物者有用的是方法能不能迁到自己的 Agent 或生成管线，不是把标题再念一遍。
研究与学习HF daily papers
huggingface.co打开原文
#62

Intern-S2-Mobius: Foundation Model with Decoupled Knowledge and Reasoning

We introduce Mobius-v0, an architecture that comprises a globally shared Memory (FFN) that stores knowledge vectors and multiple Reasoners (Self-Attn) that iteratively achieve compositional reasoning. Using hidden states

推荐理由「Intern-S2-Mobius: Foundation Mod…」对造物者有用的是方法能不能迁到自己的 Agent 或生成管线，不是把标题再念一遍。
研究与学习HF daily papers
huggingface.co打开原文
#64

Marionette: Predicting World States, Rendering Geometry, Painting Appearance

Interactive game world models typically autoregress visual observations directly in pixel or latent space, forcing structured properties such as pose, geometry, and occlusion to be implicitly maintained by the same gener

推荐理由「Marionette: Predicting World Sta…」对造物者有用的是方法能不能迁到自己的 Agent 或生成管线，不是把标题再念一遍。
研究与学习HF daily papers
huggingface.co打开原文
#65

DFM Mimir v1: An Open HRM Delivering Frontier Performance at 1B Parameters Using Only Permissible Post-Training Data

Current large language model development relies on massive, often non-permissible datasets, creating a high barrier for researchers committed to open-source and ethically sourced data. We introduce Mimir v1, a 1-billion-

推荐理由「DFM Mimir v1: An Open HRM Delive…」对造物者有用的是方法能不能迁到自己的 Agent 或生成管线，不是把标题再念一遍。
研究与学习HF daily papers
huggingface.co打开原文
#70

How Do Agents Fail on AutoResearch: End-to-End Diagnostic Evaluation on 100 Real-World Frontier Research Tasks

AI has long assisted scientific research, but the rapid advance of LLMs and agentic scaffolds is reshaping the landscape; a single system can now carry whole-stage research from an initial hypothesis all the way to final

推荐理由「How Do Agents Fail on AutoResear…」偏Agent，对造物者有用的是方法能不能迁到自己的 Agent 或生成管线，不是把标题再念一遍。
研究与学习arXiv
arxiv.org打开原文
#71

SkillComposer: Learning Reusable Skills for Natural-Language Robot Programming

Natural-language interfaces can lower the barrier to programming robots, but existing systems struggle when users request complex tasks. While large language models (LLMs) perform well with simple commands, they often st

推荐理由「SkillComposer: Learning Reusable…」偏机器人，对造物者有用的是方法能不能迁到自己的 Agent 或生成管线，不是把标题再念一遍。
研究与学习arXiv
arxiv.org打开原文
#74

Skill Blocks: How Should an Agent Load Its Skill? A Caching-Correct Comparison of Pre-load, On-Demand Tool-Loading, Progressive Disclosure,

Agent skills are often injected in full on every request, increasing token cost. We compare four content-preserving loading methods: Full, Skill Block, Reference, and Hybrid. Across SearchQA, SpreadsheetBench, ALFWorld,

推荐理由「Skill Blocks: How Should an Agen…」偏Agent，对造物者有用的是方法能不能迁到自己的 Agent 或生成管线，不是把标题再念一遍。
研究与学习arXiv
arxiv.org打开原文
#75

LLMs Can Predict Failure Risk, But Struggle to Predict Which Collaboration Protocol Pays Off: Cost-Aware Protocol Routing Across Reasoning T

Multi-agent large language model (LLM) systems can improve reasoning by spending more computation, but deployment requires deciding when extra collaboration is worth its cost. We isolate this decision by running every pr

推荐理由「LLMs Can Predict Failure Risk, B…」偏Agent，对造物者有用的是方法能不能迁到自己的 Agent 或生成管线，不是把标题再念一遍。
研究与学习arXiv
arxiv.org打开原文
#76

DarwinX: Evolving Agent Harnesses Through Natural Selection

An LLM agent's capability depends not only on model weights but on its harness: prompts, tools, skills, and control flow. Self-improvement loops already edit harnesses, yet single-lineage search is path-dependent and loc

推荐理由「DarwinX: Evolving Agent Harnesse…」偏Agent，对造物者有用的是方法能不能迁到自己的 Agent 或生成管线，不是把标题再念一遍。
研究与学习HF daily papers
huggingface.co打开原文
#77

When Is an Agent Evaluation Over? Outcome Finality and Cross-Unit Separation

Agent evaluations commonly score the state observed when a run stops and count the run as one trial. Interpreting that score as a final result from a separate trial requires outcome finality and cross-unit separation. Ou

推荐理由「When Is an Agent Evaluation Over…」偏Agent，对造物者有用的是方法能不能迁到自己的 Agent 或生成管线，不是把标题再念一遍。
研究与学习arXiv
arxiv.org打开原文
#79

Intern-S2-Preview: Scientific Agentic Foundation Model

Scientific discovery increasingly requires AI systems that can reason over scientific evidence of heterogeneous modalities, interact with scientific tools and environments, and sustain progress across long task horizons.

推荐理由「Intern-S2-Preview: Scientific Ag…」偏Agent，对造物者有用的是方法能不能迁到自己的 Agent 或生成管线，不是把标题再念一遍。
研究与学习HF daily papers
huggingface.co打开原文
#80

AutoDesign: Meta-Harness Optimization for Long-Horizon Agentic Design

Transforming multimodal sources into condensed and structured media outputs can be fundamentally conceptualized as a long-horizon agentic process centered on a model-harness system. While an ideal harness system should a

推荐理由「AutoDesign: Meta-Harness Optimiz…」偏多模态，对造物者有用的是方法能不能迁到自己的 Agent 或生成管线，不是把标题再念一遍。
研究与学习HF daily papers
huggingface.co打开原文