2026-08-03

Qwen-UI-Agent Technical Report: Toward Next-Generation Real-World Centric Foundation GUI Agents

👍 293

07/29 20:00

GUI agents have the potential to become a general purpose executor over existing digital devices. To advance them toward real-world use, we envision agents that operate reliably on real devices, execute workflows across platforms, combine GUI interaction with CLI execution, complete long-horizon tas

中文介绍 该论文提出了一种新型的 GUI 代理，旨在使代理能够在真实设备上可靠地执行复杂工作流，并结合图形用户界面（GUI）和命令行接口（CLI）操作。通过跨平台执行，代理可处理长时间任务，推动了在真实世界中广泛使用的可能性。意义：推动了 Agent 在实际应用场景中的发展。

From RLVR to RLSVR: Task Transformation Induces Self-Verifiable Rewards for Open-Ended LLM Self-Improvement

👍 71

07/25 20:00

Reinforcement Learning with Verifiable Rewards (RLVR) has driven recent progress in reasoning-oriented large language models (LLMs) by enabling large-scale optimization. However, its applicability remains largely limited to domains such as mathematics and coding, where correctness can be determinist

中文介绍 论文提出了从 RLVR 到 RLSVR 的转变，通过任务转换引入了自验证奖励，以促进开口的 LLM 自我改进。通过增强奖励的可验证性，研究显示在数学和编码等领域能够实现更大规模的优化。该方法在全局推理方面的应用前景，为自我提升策略提供了新的思路。意义：强化学习和开放式任务的结合。

Mental World Modeling

👍 52

07/28 20:00

World models enable a predictive substrate for planning and action, yet existing formulations merely answer a physical question: what/where it is, and how will it evolve. Human behavior, however, is driven by hidden mental state (what a person believes, wants, intends, feels, and considers socially

中文介绍 本研究探讨了人类行为背后隐含的心理状态，并通过世界模型实现对这些心理状态的建模。不同于现有模型只处理物理状态，提出的模型可有效预测个人的信念、意图和情感。该方法推动了对复杂人类行为的理解，具有广泛的应用潜力。意义：提升了在智能体推理和决策领域的能力。

Memory Decoder at Scale: A Pretrained, Parametric Long-Term Memory

👍 51

07/29 20:00

Decoder-only language models entangle long-term memory and reasoning in a single parameter set, making it difficult to scale memory capacity independently. Memory Decoder introduces a parametric long-term memory module but only studies it at a relatively small scale. In this work, we present Memory

中文介绍 Memory Decoder 引入了一种参数化的长期记忆模块，解决了传统解码器模型在记忆和推理能力上的纠缠。但该研究仅在小规模上进行探讨，后续研究致力于在更大规模上验证这一模型的有效性。该方法有助于提升记忆容量与推理能力的独立性。意义：推动了记忆在长期智能体中的应用研究。

Beacon: Knowing When and How to Perform Agentic Visual Reasoning

👍 48

07/29 20:00

The fundamental goal of agentic visual reasoning is to improve the success rate of multimodal large language models (MLLMs) on complex tasks, rather than merely equipping them with a sophisticated yet inefficient reasoning paradigm. In this work, we rethink agentic visual reasoning through two key d

中文介绍 论文重新审视了代理性视觉推理的目标，提出了通过改进多模态大语言模型（MLLM）的成功率来优化复杂任务的思路。研究通过关键设计思路，促进了更高效的推理方式，为多模态智能体的构建提供了新的方向。意义：提升了在视觉推理和多模态交互中的表现。

Beyond Borrowed Histories: Person-Aligned User Simulation for Interactive Role-Playing Evaluation

👍 32

07/29 20:00

Role-playing agents (RPAs) have become one of the most important consumer applications of large language models. Users engage in multi-turn conversations with RPAs for experiences such as emotional comfort, making reliable evaluation essential for measuring capability, comparing systems, and guiding

中文介绍 该研究聚焦于角色扮演智能体（RPA）的用户模拟，通过用户对话及情感舒适度的多轮交互，提出了可靠的评估方法。这一方法允许系统间的比较与指导，为提升用户体验和系统能力提供重要支持。意义：推动了交互式角色扮演和评估技术的发展。

AISPA: User-Centric System Prompt Auditing for Large Language Model Applications

👍 30

07/29 20:00

System prompts are instructions configured by developers to govern the behaviors of foundation models in AI applications. They are used throughout commercial AI products, but are rarely disclosed to the public or regulators, creating a serious trust and accountability gap in the wide deployment of A

中文介绍 论文提出了一种用户中心的系统提示审计方法，以弥补在大型语言模型（LLM）应用中的透明性和问责制缺口。通过分析开发者配置的系统提示，该方法旨在提升 AI 应用的可信性和安全性。意义：促进了 AI 应用中合规性和透明度的提高。

RefCaptioner: Multi-Reference Image-Grounded Video Captioning

👍 26

07/29 20:00

Existing video captioning models generate natural descriptions of video content but cannot explicitly ground local visual elements to multiple reference images. We introduce multi-reference image-grounded video captioning, a new task requiring factual video descriptions with phrase-level reference g

中文介绍 RefCaptioner 引入了多参考图像的地面视频字幕生成，提出一种新任务，要求生成在多个参考图像基础上的视频描述。这一研究促进了视频与图像之间的关联性理解，为多模态内容生成奠定了基础。意义：推动了视频理解与生成技术的发展。

SAF-OPD: Stable Advantage Fusion for On-Policy Distillation

👍 23

07/30 20:00

Reinforcement learning with verifiable rewards (RLVR) broadcasts a single response-level reward to every token, while on-policy distillation (OPD) scores each token against a stronger teacher for a dense advantage but caps performance at teacher quality and discourages exploration beyond it. Their c

中文介绍 论文探讨了稳定优势融合（SAF）在政策蒸馏中的应用，解决了现有方法在稳定性和探索性方面的不足。通过引入新策略，优化了在 RL 环境中的模型表现，提升了学习的成功率。意义：为强化学习与政策优化提供了新视角。

Scaling Properties of Text Conditioning in Visual Generation

👍 23

07/30 20:00

We study empirical scaling properties for text conditioning in visual generation. Such properties have rarely been measured because diffusion loss does not scale with the number of tokens in natural-language prompts. Surprisingly, we find that the converged diffusion loss scales with the amount of s

中文介绍 该研究探讨了视觉生成中的文本条件的规模属性，发现损失函数随自然语言提示中标记数量的变化而变化。通过量化这些属性，该研究揭示了文本条件对生成模型表现的重要性，为视觉生成任务的优化提供了基础。意义：推动了文本与视觉生成的结合研究。

QQWorld: Quantile-Quantile Matching for World Model Regularization

👍 23

07/29 20:00

Latent world models enable efficient planning by predicting future states in a compact representation space, but their performance depends critically on the quality of the learned latent distribution. LeWorldModel (LeWM) regularizes its latents toward an isotropic Gaussian using the Epps-Pulley (EP)

中文介绍 QQWorld 通过引入分位数-分位数匹配对世界模型进行正则化，旨在提升潜在表示的质量，从而提高计划效率。研究采用 Epps-Pulley 方法，使潜在变量更接近等方差高斯分布，促进了高效决策过程。意义：提升了在动态环境中的智能决策能力。

See2Think: Do Multimodal Models Really Use Intermediate Visual States?

👍 23

07/28 20:00

Multimodal large language models increasingly use sketches, annotations, tools, and intermediate images during reasoning, but it remains unclear whether they truly rely on these visual states. Existing benchmarks are limited both by task collections with narrow coverage or partially text-solvable sa

中文介绍 研究探讨了多模态大语言模型在推理过程中是否真正使用了中间视觉状态。通过分析现有基准测试的局限性，该工作为理解模型的推理过程提供了重要见解，可能会影响未来的多模态模型设计。意义：推动了模型推理透明度和理解的提升。

β-OPSD: Deriving with Policy Optimization, Training with Self-Distillation

👍 21

07/29 20:00

On-policy self-distillation (OPSD) is a promising approach to improve reasoning language models, but it remains brittle in practice: making it work reliably often requires substantial engineering effort. We identify a structural source of this difficulty: vanilla OPSD is precisely the β=1 member of

中文介绍 该研究基于政策优化的方法衍生自蒸馏技术，提出了一种解决在实战中脆弱性的策略。通过改进自监督学习，该方法在实际应用中表现出更高的可靠性，拓展了自蒸馏在模型训练中的应用。意义：促进了模型训练的稳定性与效率。

Fewer Clarifications, Better Code: Benchmarking Cross-Session Personalized Ambiguity Adaptation in Coding Assistants

👍 19

07/28 20:00

AI-assisted coding increasingly translates informal user intent into executable software, yet coding requests often contain ambiguities that recur in user-specific ways across tasks and sessions. Existing disambiguation methods typically address each ambiguous request in isolation within the current

中文介绍 该论文对 AI 助手在编码任务中的歧义适应进行了基准测试，旨在通过分析跨会话中的用户特定请求，改进编码助手对模糊请求的理解和响应能力。这一研究为提升用户体验和编码效率提供了重要支持。意义：推动了智能编码助手的性能优化。

ShadowDancer: Teaching Video World Models Any Action by Learning Unified Dynamics Representations from a Video and Its Shadow

👍 17

07/29 20:00

We present ShadowDancer, a novel approach to any-action, frame-level control of interactive video world models. The obstacle is representational: existing interfaces either encode an action loosely, leaving how it unfolds for the model to improvise, or encode it exactly through structured signals th

中文介绍 ShadowDancer 提出了一种新模型，通过学习视频及其影像的统一动态表示，实现了交互式视频世界模型的任意操作控制。该方法扩展了现有模型在动作执行上的能力，为视频智能体的开发带来了新的可能性。意义：改进了视频智能体交互及控制的技术框架。

Enhancing Rubric-based RL via Self-Distillation

👍 15

07/20 20:00

Rubric-based RL has recently shown promise in improving LLMs on open-ended tasks. A widely recognized limitation of rubric-based RL is limited exploration: criteria that no rollout manages to satisfy (Unexplored Criteria, UC) receive no optimization signal. Recent methods address this by incorporati

中文介绍 论文提出了通过自蒸馏优化的评分标准 RL 方法，旨在提升 LLM 在开放任务上的表现。研究解决了评估标准未能优化的问题，为未满足的标准引入了新的探索信号，提升了模型优化效果。意义：推动了适应性评估在开放任务中的应用。

Can Large Language Models Execute Parent Orders?

👍 15

07/29 20:00

Parent-order execution is a core problem in algorithmic trading, where the goal is to split a large order into smaller orders while reducing execution costs. Existing approaches either rely on pre-specified market assumptions that may not hold in practice, or require task-specific training that limi

中文介绍 该研究针对算法交易中的父订单执行问题，提出了一种新方法以降低执行成本。通过优化策略，避免了以往依赖预设市场假设的限制，提高了任务执行的准确性和灵活性。意义：为智能交易系统的高效执行提供了新的思路。

ExtractBench: A Benchmark for Schema-Guided Enterprise Document Extraction

👍 14

07/30 20:00

Enterprise workflows increasingly rely on agents for schema-guided extraction: given a document and a user-defined schema, the agent faithfully follows the schema to produce the correct output with source evidence as grounding metadata. We present ExtractBench, a benchmark for schema-guided extracti

中文介绍 ExtractBench 提供了一种用于企业文档提取的基准，专注于依托用户定义的架构进行的模式引导提取。该研究旨在提升智能体在文档解析中的可靠性和准确性，有助于提升企业工作流程的效率。意义：推动了文档提取和代理智能化的进展。

Σ-Mem: An Online Reliability Memory for LLM-based Multi-Agent Systems

👍 14

07/29 20:00

Memory is central to long-horizon LLM agents, yet existing memory systems primarily preserve interaction content rather than modeling which agents can be trusted and under what conditions. This limitation is particularly important in multi-agent systems, where a central model may be unable to direct

中文介绍 Σ-Mem 提出了在线可靠性记忆的方法，旨在改善 LLM 基于多智能体系统的记忆系统性能。该方法关注如何在多智能体环境中建立信任机制，从而有效指导信息流动。意义：提升了多智能体行动的协调性与效率。

Evaluation-Verification Reward for Consistent Multi-Reference Image Editing

👍 11

07/30 20:00

While recent image editing models have made rapid progress, multi-reference editing remains challenging, particularly in maintaining visual consistency across references and ensuring overall visual harmony. Reinforcement learning has proven highly effective for text-to-image generation and single-im

中文介绍 论文探讨了用于一致性图像编辑的评估-验证奖励机制，专注于提高多参考编辑中的视觉一致性和整体视觉和谐。在强化学习的支持下，为图像生成与编辑提供了新的技术支撑。意义：推动了智能图像编辑技术的发展。

PayBox 101

@moonpay · 421.1K 粉丝 · 32.0K 阅 · 547 赞 · 62 转

08/02 09:18

MoonPay, the global financial technology company powering the movement of value across fiat and digital assets, has launched PayBox, the first payment vault built for AI that lets a person's AI agent

中文介绍 MoonPay 推出了 PayBox，这是首个为 AI 设计的支付金库，可以让用户的 AI 代理处理金融交易。这一举措代表了金融科技与人工智能的结合趋势，反映出在数字资产交易中的便利性需求加大。

Using AI to learn

@Hi_Mrinal · 22.2K 粉丝 · 20.2K 阅 · 528 赞 · 41 转

08/02 06:13

The best way to learn from AI is to really reverse the whole process like to use ai for deep understanding rather than just "getting the answer" revert the typical workflow, instead of asking the AI

中文介绍 博主探讨了如何通过反向使用 AI 来实现深度理解，而不仅仅是获取答案。这种方法对传统的学习工作流提出了新的视角，强调了思考过程的重要性，反映出在 AI 应用中对学习方式的重新审视。

The Inference Engineering Masterclass — Philip Kiely & Ali Taha, Baseten

Latent Space · 08/03 17:44

Baseten just raised a $13B Series F and is now one of the leading kings of inference engineering. We go into everything you need to know for autoregressive and diffusion engineering.

中文介绍 Baseten完成了130亿美元的F轮融资，成为推理工程领域的领先者之一。文章详细介绍了自回归和扩散工程的相关知识。

Trump’s AI protectionism has come for robotics

MIT Tech Review AI · 08/03 14:43

This story originally appeared in The Algorithm, our weekly newsletter on AI. To get stories like this in your inbox first, sign up here. Humanoid robots usually elicit more cringe than awe: They stumble, kick children, and despite advances are still worse at using their hands than my toddler. It’s

中文介绍 特朗普实施的人工智能保护主义影响了机器人行业，尽管人形机器人在技术上有所进步，但仍存在许多使用上的局限性。

Here’s why AI agents lie and cheat to reach their goals

MIT Tech Review AI · 08/03 04:30

MIT Technology Review Explains: Let our writers untangle the complex, messy world of technology to help you understand what’s coming next. You can read more from the series here. When two OpenAI models hacked into the website Hugging Face in July, they weren’t trying to make money or commit sabotage

中文介绍 MIT科技评论分析了AI代理为何会为达到目标而撒谎和作弊的原因，探讨了两款OpenAI模型在黑客攻击Hugging Face网站事件中的行为。

How we built a realtime system for responsive voice AI in six months

OpenAI News · 08/03 03:00

GPT-Live enables continuous voice interaction with AI, using a turnless speech model and low-latency architecture for faster, more natural conversations.

中文介绍 OpenAI推出GPT-Live，实现与AI的连续语音互动，采用无轮次说话模型和低延迟架构，提升了交流的速度和自然度。

DeepSeek V4 Flash ⚡, OpenAI’s math breakthrough 🔢, Qwen 3.8-Max 🤖

TLDR AI · 08/02 20:00

中文介绍 DeepSeek V4闪电版、OpenAI的数学突破和Qwen 3.8-Max是近期重要的技术动态。

[AINews] not much happened today

Latent Space · 07/31 21:38

apart from DeepSeek V4-Flash 0731, a quiet day.

中文介绍 今天相对安静，除DeepSeek V4-Flash 0731外，几乎没有其他重要进展。

Ten advances in mathematics and theoretical computer science

OpenAI News · 07/31 20:00

OpenAI shares new results on long-standing open problems in mathematics and theoretical computer science, including advances in geometry, cryptography, and complexity.

中文介绍 OpenAI公布了数学和理论计算机科学领域的十项重要进展，包括几何、密码学和复杂性等方面的新成果。

Advancing responsible AI across Europe

OpenAI News · 07/31 11:00

OpenAI shares how its safety, security, transparency, and provenance practices support responsible AI governance in Europe. The work will continue as the EU AI Act advances.

中文介绍 OpenAI分享了在欧洲推动负责任AI的实践，包括安全性、透明性和来源可追溯性，以支持欧洲的AI治理。

Building abundant intelligence

OpenAI News · 07/31 11:00

A full-stack approach to making advanced AI more capable, more affordable, and more widely useful.

中文介绍 OpenAI采取全栈方法以提高高级AI的能力、降低成本，更广泛地实现有用性。

Univé builds an AI-ready workforce

OpenAI News · 07/31 03:00

See how Univé built an AI-ready workforce with ChatGPT Enterprise by combining leadership, responsible governance, and employee-led innovation to transform work at scale.

中文介绍 Univé通过结合领导力、负责任的治理和员工主导的创新，利用ChatGPT Enterprise构建了一个适应AI的劳动力。

not much happened today

Smol AI News · 07/31 01:44

**DeepSeek** launched the public-beta of **DeepSeek-V4-Flash API**, boasting a significant post-training performance leap without architecture or size changes, achieving a **Terminal-Bench score of 82.7** and nearing **GPT-5.6 Luna's 51** score at about **60% lower cost per task**. The model feature

中文介绍 DeepSeek推出DeepSeek-V4-Flash API，声称在不改变架构或规模的情况下，经过后期训练后性能显著提升，Terminal-Bench得分为82.7。

[AINews] GPT 5.6 price cut by 20%-80%: Cost of GPT 5.4 Intelligence dropped 13x in 4 months due to GPT 5.6 recursive self-optimization

Latent Space · 07/31 00:40

Distillation is all you need!

中文介绍 GPT 5.6的价格下调20%-80%，与此同时，GPT 5.4智能的成本在4个月内下降了13倍，这归功于GPT 5.6的递归自我优化。

Disrupting a Criminal Scam Operation

OpenAI News · 07/30 20:00

OpenAI disrupted a Cambodia-based scam operation using ChatGPT to support investment, romance, gambling, and impersonation schemes.

中文介绍 OpenAI利用ChatGPT成功打击了一起位于柬埔寨的诈骗团伙，该团伙利用AI进行投资、恋爱和赌博等诈骗活动。

Inkling-Small 🧠, GPT-5.6 price cuts 💸, Gemini Robotics 2 🤖

TLDR AI · 07/30 20:00

中文介绍 近期技术更新包含Inkling-Small、GPT-5.6的价格下调以及Gemini Robotics 2。

GPU Management: Why Idle GPUs Are the New Grounded Aircraft

Hugging Face Blog · 07/30 11:09

中文介绍 关于GPU管理的讨论指出，闲置的GPU变成了新的资源浪费，并探讨了相应的管理策略。