2026-08-13

AI4AI at Test-Time: Strong-to-Weak Capability Transfer via Harnesses

👍 40

08/12 08:00

Recent work on distillation transfers the capabilities of large models to smaller ones often by updating the latter's parameters, through teacher forcing, on-policy distillation, and related training-time methods. In this paper, we ask whether such transfer can instead occur at test time. We study s

From Synthesis to Removal: Physics-Grounded Reflection Simulation and Diffusion-Based Video Dereflection

👍 3

08/12 08:00

Videos captured through glass often contain reflections that degrade visual quality and interfere with downstream vision tasks. Although single-image reflection removal has been extensively studied, video reflection removal remains largely underexplored due to the lack of paired video data, temporal

MBA: Multimodal Benchmark and Agents for Real-World Business Ideation

👍 1

08/12 08:00

Agentic systems powered by large language models (LLMs) have opened new opportunities for business ideation. Yet existing approaches remain confined to a text-only paradigm, despite the inherently multimodal nature of real-world contexts. We thus introduce MBA-Bench, the first multimodal benchmark f

AVA-Encoder: Towards Agent-Native Video Representation Learning

👍 0

08/12 08:00

Creative agents still lack an effective way to learn from high-quality human films, limiting their ability to produce cinematic-grade videos. A key challenge is the absence of a structured video representation that is both faithful to film content and directly usable for agentic reasoning and manipu

Spark-to-Paper: End-to-End Research Paper Generation as a Composable Skill

👍 9

08/12 08:00

Turning a research idea into a complete paper requires more than text generation: the system must retrieve literature, design and execute experiments, revise claims according to evidence, produce publication-ready figures, and maintain consistency across a long generation process. We present Spark-t

ToolHazard: Scaling Adversarial Environments for Security Evaluation and Alignment of LLM-based Agents

👍 4

08/12 08:00

Large language model (LLM) agents integrated with external tools are vulnerable to indirect prompt injections embedded in environmental states. However, existing studies largely rely on manually implemented or reused environments, stochastic LLM-based tool simulation, and predefined injection locati

Agent Safety Should Be a Runtime Contract

👍 1

08/11 08:00

The dominant paradigm treats AI safety as a property to be instilled during model training via RLHF, DPO, or Constitutional AI. We argue this is structurally insufficient for autonomous agents that execute code, mutate files, send messages, and modify databases. Agent safety should be a runtime cont

InSight-doc: Agentic Visual Perception for Long-Document Understanding

👍 10

08/11 08:00

Long-document understanding often requires reasoning over many visually rich pages, making inference costly and prone to context rot. In this work, we propose InSight-doc, an agentic visual perception framework that treats visual resolution as an adaptive reasoning-time resource. InSight-doc starts

DistilVDR: A Compact End-to-End Visual Document Retriever via Dual-Student Distillation

👍 10

08/11 08:00

Visual document retrieval (VDR) is dominated by multi-billion-parameter models that are slow to index at full corpus scale and expensive to serve. Prior compression routes either train a smaller multi-vector encoder from scratch or distil only the query side; neither yields a compact single-vector r

Reference-Free Post-Training of Open Large Language Models for Multilingual Machine Translation

👍 10

08/11 08:00

We study reference-free post-training for multilingual machine translation with open large language models. Starting from the supervised-finetuned MiLMMT-46-v0.1 models, we apply Group Relative Policy Optimization (GRPO) with a reward that averages two reference-free quality estimation models and is

ComBodied Agents: a New Paradigm of Human-Centric Agentic AI

👍 176

08/11 08:00

After an older adult misses a medication dose, a software agent can send another reminder and an embodied agent can bring the medication. Yet neither explains whether the person forgot, is confused, has side effects, or deliberately refused, nor what support is appropriate. This reveals a structural

SkillZip: Evaluation-Free Skill Compression for Self-Evolving Agents by Discovering Reusable Structure

👍 11

08/11 08:00

Self-evolving agents accumulate reusable skills by appending successful procedures and failure fixes. Over time, the same requirement is often restated in several branches, examples, and warnings, while common action sequences are copied rather than reused. The resulting skill becomes expensive to i

DSAgentBench: Can Agents Automate End-to-End Data-Science Workflows in Real Computer Environments?

👍 7

08/11 08:00

Real-world data science involves long-horizon workflows that span data wrangling, exploration, modeling, visualization, and validation, and require coordinated use of tools such as notebooks, IDEs, terminals, browsers, and databases within real operating environments. Yet existing benchmarks lack re

SPIEval: Evaluating Large Language Models as Mobile Assistants over Scattered Personal Information

👍 10

08/11 08:00

Large language models (LLMs) are increasingly deployed as mobile assistants, where a key challenge is leveraging personal information scattered across multiple applications (apps) to complete user instructions. However, due to the lack of dedicated benchmarks, their capabilities remain poorly unders

VibeLifeBench: Can Your Life Agent Be Proactive and Persistent in a Living World?

👍 15

08/11 08:00

Large language model (LLM) agents are increasingly deployed as personal assistants. Existing evaluations, however, mostly use short, self-contained requests in static environments. Everyday life assistance is different. A task runs for weeks rather than minutes. The world keeps changing while the ag

Power law graph attention: exact generalization of scaled dot-product attention, empirical collapse at inference

👍 6

08/10 08:00

The Large Language Model from Power Law Decoder Representations (PLDR-LLM) and its attention, Power Law Graph Attention (PLGA), replace the fixed bilinear form of scaled dot-product attention (SDPA) with a learned, input-generated bilinear operator G_{LM}, built from a positive tensor A_{LM} by elem

Decoding-Level Taboo: A Diagnostic Stress Test for LLM Robustness

👍 12

08/10 08:00

Large language model evaluations typically focus on performance under nominal conditions, creating an illusion of capability where models comfortably walk a narrow, highly optimized generation corridor. In real-world deployments, however, complex system prompts, safety guardrails, and structural con

Co-Evolution in Agentic Systems: Toward Self-Directed Evolution Beyond Human Design

👍 121

08/10 08:00

Agentic systems are increasingly expected to improve after deployment, yet single-entity self-evolution is often bounded by a static learning context, such as fixed tasks and feedback. This survey focuses on co-evolution in agentic systems, a multi-component form of self-evolution in which multiple

360CityArena: A Realistic Virtual Urban Navigation Benchmark for Embodied Agents

👍 9

08/09 08:00

We present 360CityArena, a benchmark for evaluating the urban exploration capabilities of embodied agents within a photorealistic environment constructed from 360-degree videos. Existing outdoor benchmarks either lack sufficient photorealism or complexity, resulting in a considerable gap from real-w

Not Worth Another Token: Marginal Value Estimation for Efficient Deep Research Agents

👍 11

08/09 08:00

Long-horizon research agents solve open-ended tasks through iterative retrieval, aggregation, and synthesis, but context grows rapidly while the marginal value of additional evidence often declines. This leads to unnecessary token cost, higher latency, and noisier inputs for final report generation.

VectraYX-Vision-1B: A Sub-2B Spanish/LATAM Cybersecurity Vision-Language Model with Structured Visual Reasoning and Native Tool Use

👍 6

08/09 08:00

We present VectraYX-Vision-1B, a sub-2B vision-language model (VLM) for Spanish/LATAM cybersecurity imagery, coupling a frozen SigLIP-so400m encoder to a 1.04B Spanish/LATAM security decoder via an MLP. To our knowledge, it is the first sub-2B VLM specialized for cyber UI (IDA, Ghidra, Wireshark, Nm

Can LLM Agents Stick to the Script? A Benchmark for Long-Horizon Consistency in Interactive Narratives

👍 14

08/08 08:00

The rapid advancement of Large Language Models (LLMs) is revolutionizing AI for Games by enabling open-ended and fluid interactive storytelling. However, existing research has largely overlooked the critical challenge of maintaining long-horizon logical consistency and narrative integrity against un

NeuPAT: Neuron-aware Plasticity Allocation Tuning for Language-Preserving MLLMs

👍 1

08/08 08:00

Multimodal expansion of large language models (LLMs) enables new perceptual capabilities but often compromises the language intelligence acquired during pretraining. In this work, we investigate this phenomenon from the perspective of internal adaptation dynamics and discover that neurons in pretrai

TSDS-Toolbox: A Toolbox for Measuring Time-Series Dataset Similarity

👍 8

08/08 08:00

The rapid advancement of artificial intelligence (AI) has significantly accelerated research in time-series analysis, particularly in forecasting, classification, and generation tasks. Recent models, especially foundation models, benefit from time-series dataset similarity due to its significant rol

CoinRAG: Contextualized Information Nugget KV Cache Reuse for Long-Context RAG

👍 4

08/07 08:00

Recent optimization studies on Retrieval-Augmented Generation (RAG) have exploited chunk-level KV cache reuse to avoid processing long retrieved contexts for higher efficiency, while significant information redundancy and noise still remain in the coarse-grained chunks. This paper optimizes the Pare

iFAN: Inference-Aware Learning for Plain Mask Transformers

👍 8

08/07 08:00

Query-based mask transformers assemble segmentation outputs through pixel-wise competition among query predictions of the final layer, yet this inference process is not explicitly optimized during training. We identify two key mismatches: the query with the highest probability-mask score does not ne

Mendel Gödel Machine: Recursive Self-Improving Coding Agents via Comparative Evolution

👍 23

08/07 08:00

Self-improving coding agents that iteratively rewrite their own source code have demonstrated impressive performance on coding tasks. However, existing solutions generally derive self-modification from a single failure trajectory at a time, overlooking rich comparative signals available in the agent

The Illusion of Visual Tool-Use: A Causal Audit of Thinking with Images

👍 1

08/06 08:00

The "thinking-with-images" paradigm equips multimodal LLMs with active visual operations such as crop-and-zoom. However, models using these operations often achieve only marginal or negative gains over direct inference at substantially higher token cost. They may also repeatedly crop irrelevant regi

JigShape: Evaluating Visual-Geometric Reasoning in VLMs through Jigsaw Puzzles

👍 7

08/04 08:00

Jigsaw puzzle solving requires jointly reasoning about visual content and geometric constraints, yet existing benchmarks use rectangular cuts that create ambiguous ground truth in texture-repeated regions. We introduce \ours{}, a benchmark with tab-and-blank interlocking pieces where geometric const

AutoWorldModel-Bench: A State-Centric Benchmark for Automated World-Model Research

👍 1

07/20 08:00

World modeling is an unsettled field: architectures, training objectives, and state representations interact in complex ways, and no single recipe dominates across environments. This makes it an ideal testbed for AI coding agents acting as autonomous researchers--a setting in which the improvement d

To FDE, or not to FDE?

@thejessezhang · 85.7K 粉丝 · 270.4K 阅 · 510 赞 · 30 转

08/11 23:24

Two-thirds of our deployment work is now done autonomously by our own product (via Duet). This post is about why that is, our philosophy on building a product + deployment motion, and how that's

中文介绍 探讨自动化部署哲学，阐述产品与部署动力的构建思路。

Agentic Code Quality

@addyosmani · 408.3K 粉丝 · 175.5K 阅 · 509 赞 · 65 转

08/12 14:36

For much of human history, we've evaluated code quality via code review: someone reads what you wrote and makes sure it's clean, thoughtful, fast, understandable, and tests well. For agents, that

中文介绍 分析代码审查在人类历史中的作用，探讨AI代理的代码质量评估方式。

How to Build a Shared AI Harness for Your Team

@VibeMarketer_ · 30.1K 粉丝 · 107.1K 阅 · 528 赞 · 29 转

08/10 21:36

I am going to show you how to turn the models, agents, skills, and automations living across your team's tools and workflows into one coordinated system. What most teams have today looks very

中文介绍 介绍构建团队共享AI工具的方法，实现团队工具和工作流程的协调系统。

NVIDIA Nemotron 3.5 Lightning and NeMo Switchyard Deliver Faster, Smarter, More Efficient Agentic AI

@nvidia · 2.6M 粉丝 · 59.5K 阅 · 548 赞 · 66 转

08/11 21:41

The new lightweight open model and routing library delivers greater control over AI, data and workflows across edge devices, PCs, workstations, data centers and the cloud. Source: NVIDIA Blog By Kari

中文介绍 NVIDIA发布Nemotron 3.5和NeMo Switchyard，提升边缘设备、PC、工作站、数据中心和云端的AI控制效率。

7 rules for self-improving agent loops every AI engineer should know

@GoogleCloudTech · 1.3M 粉丝 · 46.8K 阅 · 505 赞 · 63 转

08/11 01:57

By @Saboo_Shubham_ and @secchi_elia When coding agents write the code, defining what's good is the real engineering job. Coding agents can now build and improve other agents. Write the instructions,

中文介绍 分享7条自我改进的AI代理循环规则，适用于AI工程师。

/show-me: compact visual representations for coding agents

@dexhorthy · 30.0K 粉丝 · 42.9K 阅 · 635 赞 · 31 转

08/12 23:59

tl;dr make your agent converse visually instead of in walls of prose. Lighter and faster than HTML, good enough for most dev-work shaped problems. Coding agents are pretty much unreadable The

中文介绍 介绍/Show-me工具，以视觉方式简化代理对话，提高开发效率。

Intro to Grok Bot

@mattyp · 44.5K 粉丝 · 33.5K 阅 · 651 赞 · 48 转

08/12 02:59

I’m always chasing better tools. Notes apps, workflows, optimizations, and now, personal agents. It’s an easy trap to fall into - always building the “custom” thing and sacrificing the work as a

中文介绍 介绍Groks Bot，强调避免过度定制，牺牲工作质量。

Grok 4.6 – A field guide

@ericzakariasson · 80.4K 粉丝 · 32.2K 阅 · 719 赞 · 52 转

08/12 23:46

Grok 4.6 is out! I've used it for a few weeks as my daily driver across the normal mix of coding and knowledge work, and built a few projects with it specifically to push on where it holds up. It's

中文介绍 发布Grok 4.6版本，分享使用经验和项目案例。

[AINews] SpaceXAI Grok 4.6 and Grok @Bot

Latent Space · 08/13 09:53

AI teammate category just had its most significant new entrant yet

中文介绍 SpaceXAI发布Grok 4.6和Grok @Bot，AI团队成员类别迎来重大新成员。

Scaling AI agents with trustworthy data

MIT Tech Review AI · 08/13 00:51

Business and technology leaders need no convincing that the time of agentic AI is here. Organizations are rapidly adopting agents, and few executives doubt the technology’s potential to transform work. But many organizations find that realizing the desired return on investment (ROI) from AI hinges o

中文介绍 企业领导者普遍认为智能体AI时代已到来，但许多组织发现实现预期投资回报面临挑战。

Introducing OlmoEarth embeddings: Custom embedding exports from OlmoEarth Studio for downstream analysis

Hugging Face Blog · 08/13 00:14

中文介绍 Hugging Face推出OlmoEarth Studio的定制嵌入导出功能，用于下游分析。

Putting sign language AI into users’ hands

DeepMind Blog · 08/12 22:01

Introducing sign-language-to-text (SL2T), our breakthrough model powering new sign language features for Deaf and hard of hearing users.

中文介绍 DeepMind推出突破性模型SL2T，为听障和听力受损用户提供新的手语功能。

LFM2.5-VL-3B for Better and Faster Vision Capabilities for the Edge

Hugging Face Blog · 08/12 22:00

中文介绍 Hugging Face推出LFM2.5-VL-3B，提升边缘视觉能力。

[AINews] How to steal a Reasoning Trace

Latent Space · 08/12 15:11

Speculative Decoding by any other name would distil as sweet

中文介绍 Latent Space探讨如何窃取推理跟踪，即Speculative Decoding。

From assistance to execution: How enterprises put AI to work

OpenAI News · 08/12 14:00

OpenAI research reveals how enterprises are adopting agentic AI, using ChatGPT and Codex, and how frontier firms are pulling ahead in AI adoption.

中文介绍 OpenAI研究显示企业如何采用智能体AI，使用ChatGPT和Codex，前沿企业AI应用领先。

How RingCentral builds AI-native work from engineering to ops

OpenAI News · 08/12 08:00

See how RingCentral uses ChatGPT Work and Codex to accelerate AI product development and centralize operational intelligence across engineering and operations.

中文介绍 RingCentral利用ChatGPT Work和Codex加速AI产品开发，集中运营智能。

Cursor’s GitHub rival 👨‍💻, OpenAI COO leaves 👋, Gemini 1B milestone 📈

TLDR AI · 08/12 08:00

中文介绍 OpenAI COO离职，Gemini 1B里程碑达成，xAI的Grok 4.6和Alibaba的Qwen3.8-Max模型更新。

🔬The BioAI Phase Shift - Matthew McPartlon & Neil Patil, Chai Discovery

Latent Space · 08/12 05:03

Pharma is suddenly paying for Bio × AI tools, and Chai is leading the pack with four deals closed this summer. Cofounder Matt McPartlon and Product leader Neil Patil explain why.

中文介绍 Chai Discovery在BioAI领域取得进展，夏末完成四笔交易。

Thinking of ACE? We Can Do It with Fewer Tokens

Hugging Face Blog · 08/11 21:37

中文介绍 Hugging Face探讨使用更少标记实现ACE的可能性。

Testing ads in ChatGPT

OpenAI News · 08/11 18:00

OpenAI begins testing ads in ChatGPT to support free access, with clear labeling, answer independence, strong privacy protections, and user control.

中文介绍 OpenAI开始在ChatGPT测试广告，以支持免费访问，确保隐私保护和用户控制。

Daybreak models are now available on AWS

OpenAI News · 08/11 18:00

OpenAI and AWS are making Daybreak cybersecurity capabilities available through Amazon Bedrock to support enterprise security workflows.

中文介绍 OpenAI与AWS合作，将Daybreak网络安全能力通过Amazon Bedrock提供给企业。

not much happened today

Smol AI News · 08/11 13:44

**xAI's Grok 4.6** advances frontier pricing and performance, scoring **61 on the Intelligence Index** and showing strong agentic results, with **Grok 4.7** already in training. **Alibaba's Qwen3.8-Max** open weights release features a **2.4T parameter model with 95B active MoE**, notable for day-0

中文介绍 xAI的Grok 4.6在智能指数上获得61分，Grok 4.7已开始训练；Alibaba的Qwen3.8-Max模型开放权重。

[AINews] Muse Glimmer and Spark: Open Weights return Personal Superintelligence promise

Latent Space · 08/11 13:16

a small win for american open models - Glimmer runs on a fits on a single RTX 3090!

中文介绍 Muse Glimmer和Spark：美国开放模型取得小胜利，Glimmer在单个RTX 3090上运行。