Knowledge catalogue
Coding agents break when models are 'almost' bug-free. But almost valid JSON is just not the same valid JSON. Fun piece here from @akshay_pachaar shows why SFT can't fix this, and how GRPO trains agai
arXiv:2606.11182v1 Announce Type: cross Abstract: In this paper, we propose EEVEE, the first multi-dataset test-time prompt learning framework for LLM agents, enabling test-time prompt learning under
arXiv:2606.09900v1 Announce Type: cross Abstract: Long-term memory is the missing layer for LLM agents: across sessions they forget, and the common workaround -- replaying the whole history into the p
A.J. Vicens / Reuters: The FBI seizes 13 domains allegedly tied to fake consulting firms that sought information from US government and military employees for suspected Chinese agents — Federal author
arXiv:2606.11045v1 Announce Type: new Abstract: Reusing a held-out benchmark adaptively should, in principle, invite overfitting. Yet benchmark-driven machine learning (ML) has produced surprisingly l
NVIDIA FLARE Auto-FL automates federated learning research by constraining agent actions through a control plane, enforcing fixed benchmark contracts, and using an experiment ledger to ensure reproduc
arXiv:2606.08875v1 Announce Type: new Abstract: Optimizing large language models (LLMs) for long-horizon caregiver agents requires balancing delayed task objectives with immediate environment dynamics
A developer used Replit's parallel agents feature to simultaneously create multiple deliverables for a travel app startup: a functional mobile application, promotional video, and investor pitch deck.
Google and Hugging Face are launching the Fast Gemma Challenge, where multiple agents will collaborate to optimize the performance and speed of Gemma 4 E4B model. The initiative aims to improve the ef
arXiv:2606.07711v1 Announce Type: cross Abstract: Memory is the key component for transforming a stateless LLM into a persistent, evolving agent through experience accumulation, long-horizon planning,
Three MLX videos dropped at WWDC: Running agents locally by @angeloskath https://www.youtube.com/watch?v=wykPErJ8M-8 Distributed inference and training by Tatiana Likhomanenko https://www.youtube.com/
Great tips. In practice, this is how it roughly looks to run agents autonomously for hours or days. /goal or /loop to keep it going. Verification is crucial here. Seeing a number of benchmarks showing
Amazon Bedrock AgentCore Runtime gives each agent session its own isolated microVM with a persistent workspace, secure tool access through Gateway, and built-in observability—so you can run Claude Cod
arXiv:2606.07074v1 Announce Type: cross Abstract: Deep research agents have demonstrated remarkable capabilities in complex information-seeking tasks, yet this power comes at a steep computational cos
This article explores how Databricks' Genie product and AI agents can automate and improve the processing of maintenance reports for renewable energy installations, specifically solar and wind facilit
This post seeks guidance on configuring local AI infrastructure on an AMD-based Linux system (7900 XTX GPU with CachyOS), specifically covering the setup of large language models via Ollama, AI agents
Cristina Criddle / Financial Times: OpenAI plans to overhaul ChatGPT in the coming weeks, turning it into a superapp with coding tools and AI agents to serve as a gateway to higher-margin products — $
arXiv:2606.05679v1 Announce Type: cross Abstract: Agents increasingly generate SQL, orchestrate pipelines, and automate data analysis on behalf of users. While recent work improves query correctness,
arXiv:2606.05784v1 Announce Type: new Abstract: We identify and formally characterize credit misassignment as a systematic failure mode of GRPO in tool-augmented multimodal search agents: its uniform
arXiv:2606.06284v1 Announce Type: new Abstract: Large language model agents increasingly rely on external tools, but larger tool menus can reduce reliability and efficiency by increasing wrong-tool ca
arXiv:2606.06055v1 Announce Type: new Abstract: Long-term memory enables language model agents to support personalized interactions, but it remains unclear when available memories warrant integration
arXiv:2606.05806v1 Announce Type: new Abstract: Existing benchmarks evaluate Tool-Integrated Reasoning (TIR) in LLMs on idealized ''happy paths'', largely overlooking real-world tool failures. We intr
arXiv:2606.05553v1 Announce Type: new Abstract: Role-playing language agents (RPLAs) should play characters whose values and behavior evolve as the story progresses, not maintain a fixed persona. Exis
arXiv:2606.05920v1 Announce Type: cross Abstract: Existing code-generation benchmarks score a single mapping from a complete prompt to a one-shot output. However, real web development is different. Us
arXiv:2606.05894v1 Announce Type: new Abstract: Long-horizon agents can archive large histories, but future answers still incur retrieval, rereading, and context costs. When retained memory misses ans
Storyboarding in Grok @imagine was pretty fun. I really like how the agents help visualize some of the important historical events. I was trying to iterate through the clothing and voices from the anc
Working with agents should feel like working with a colleague. You should be able “speak to” them not just with text chats, but by gesturing at a screen together, talking live, etc. With Design Mode,
arXiv:2510.15416v2 Announce Type: replace Abstract: We investigate a framework in which LoRA adapters are treated as callable tools that a base language model can dynamically select and invoke. We hyp
Zoe Kleinman / BBC: An interview with Naomi Gleit, Meta's head of product who joined the company 20 years ago, on Zuckerberg's “unfair” reputation, AI agents' capabilities, more — When Naomi Gleit joi
arXiv:2606.04141v1 Announce Type: cross Abstract: LLM agents often place sensitive credentials in the same context window as untrusted retrieved content, creating a direct path for indirect prompt inj
Nemotron 3.5 ASR is built for streaming multilingual speech recognition and voice agents. One 0.6B checkpoint. 40 language-locales. Sub-100ms latency. Cache-aware FastConformer carries context forward
This post describes a user's account suspension by OpenAI shortly after making a payment, resulting in loss of access to multiple Codex agents and client-dependent income, with no explanation provided
Shared my first trace from @NanoClaw_AI to @huggingface yesterday. Very cool! By default, all agents should store their traces on HF (in private) so that you can keep a history of them, analyze them,.
arXiv:2606.03239v1 Announce Type: new Abstract: LLM-based search agents are trained predominantly with outcome-only reward, leaving the search process itself unsupervised. This signal degenerates on o
arXiv:2606.03678v1 Announce Type: new Abstract: Generating safety-critical scenarios is essential for validating and improving autonomous driving systems, yet it inherently requires maximizing adversa
Pinecone Nexus has integrated with Microsoft OneLake, enabling AI agents to access and work directly with enterprise data stored in OneLake without requiring data movement or copying. This integration
arXiv:2510.16392v3 Announce Type: replace Abstract: Personalized and continuous interactions are critical for LLM-based conversational agents, yet finite context windows and static parametric memory h
the best agents aren't just built with the best models: they're built with harnesses purpose-built for the task at hand here's a guide on how to build a harness that's really good at feeding the model
Uber reportedly now caps coding agents at $1,500/month per employee per tool - seems sensible to me, but it's also an interesting hint at the value Uber thinks these tools are providing https://simonw
Build on-device personal AI agents on Windows PCs with new tools from NVIDIA and Microsoft, including secure sandboxing, faster local inference, multi-GPU support, and RTX acceleration for Windows AI
Cisco Systems Inc. today unveiled a set of collaboration and customer experience products that transform its Webex videoconferencing platform into what executives describe as an intelligent operating
Conductor's parallel coding agents were local-only, but now they can run remotely on Vercel. 'Our customers can't tell the difference because Vercel's Sandboxes are so fast.' https://vercel.com/blog/h
arXiv:2606.02204v1 Announce Type: new Abstract: Large language model agents achieve strong performance on text-based benchmarks but incur prohibitive inference costs, motivating the use of compact neu
arXiv:2606.00476v1 Announce Type: new Abstract: Do LLM agents act on the reasoning they state? This question of process fidelity is central to using LLMs in social simulation, yet it is hard to measur
Holo3.1 is a computer use agent developed by Hugging Face that enables fast, local execution of tasks on computing systems without requiring cloud infrastructure. The model is designed to interpret an
Research demonstrates that LLM-based agents can generate functionally correct patches that pass all tests while still containing security vulnerabilities, challenging the assumption that test-passing
arXiv:2606.01848v1 Announce Type: new Abstract: Search-and-rescue (SAR) requires embodied agents to explore unfamiliar environments under multimodal uncertainty, perform multi-stage interactions, and
This broadcast discusses techniques for improving performance and speed in Replit by utilizing parallel agents or concurrent processing methods. The content likely covers how developers can leverage p
arXiv:2605.30590v1 Announce Type: cross Abstract: Two clinical AI systems can score nearly identically on coverage-based rubrics yet behave radically differently when their patient inputs change: one
arXiv:2602.10809v2 Announce Type: replace Abstract: Existing multimodal retrieval systems excel at semantic matching but implicitly assume that query-image relevance can be measured in isolation. This
arXiv:2605.30771v1 Announce Type: new Abstract: AI agents that persist across sessions need memory they can retrieve, audit, update, and erase. Existing memory systems often collapse source evidence,
arXiv:2511.04393v2 Announce Type: replace Abstract: Large language models (LLMs) are increasingly deployed as 'agents' for decision-making (DM) in interactive and dynamic environments. Yet, since they
In this post, we use a lakehouse data agent to demonstrate how you can use Policy for deterministic access control and Lambda interceptors for dynamic validation. We then show how to combine Lambda in
arXiv:2605.31408v1 Announce Type: cross Abstract: Skill documents provide procedural knowledge to large-language-model agents at inference time. This article studies whether the presentation granulari
The gains are most obvious in coding and operational areas. Even conservative organizations are adopting coding agents fast. That doesn't mean AI adoption doesn't come with huge challenges and cost co
/goal and other fully automated AI agents are cool, but not a great model for the future of work with people. Instead you want your AI to know when to ask you GOOD questions, maybe because it is stuck
Deep Agents v0.6 makes harness profiles a first-class abstraction. Now, you can get production-grade performance from models like @Kimi_Moonshot, @Alibaba_Qwen, and @DeepSeek_ai at 20x+ lower cost tha
arXiv:2508.05614v2 Announce Type: replace-cross Abstract: LLM agents achieve 85-96% success on tasks where instructions fully specify the action, but drop to 29-53% when action feasibility depends on
I watched the LangChain keynote expecting product releases. Got something better — the clearest breakdown I've heard of why most agents never make it past demo. One thesis lands hard. Here's what sepa
arXiv:2511.22884v2 Announce Type: replace Abstract: Data analysis has become an indispensable part of scientific research. To discover the latent knowledge and insights hidden within massive datasets,