All entries

We have greatly expanded the surface area for plugins over the last few weeks, and now you can extend LLM Inference Providers and Gateway Channels with plugins. Want a custom implementation of your fa

Harrison Chase expresses support for expanding AI benchmarking efforts and highlights appreciation for work being done by Harvey (likely Harvey AI, a legal AI company). The post indicates enthusiasm f

We use previous generations of Composer to train future ones. Our autoinstall system has earlier Composer models set up dev environments for RL training. That way, the next generation can focus on lea

we're continuing to see clear examples where a model's harness is a major determinant of overall performance. with the same model, running on same task, it's easy to observe very different scores depe

We're doubling down on keeping your Apps secure 🔒 Starting today, all Replit builders- free and paid- can publish their apps privately, making them accessible only to the people you choose. When publi

We're launching the agentic robotics app store today. Let's democratize AI robotics for all! 300+ apps shipped. 10,000 robots in the wild. It used to take weeks from a robotics engineer to build apps,

We're winding back our peak hours limit reduction and doubling 5 hour limits. Excited to partner with SpaceX to bring you more compute and we'll keep pushing to bring you the best coding agent in the

We’ve agreed to a partnership with @SpaceX that will substantially increase our compute capacity. This, along with our other recent compute deals, means that we’ve been able to increase our usage limi

We’ve been working closely with the @harvey team on the launch of the Legal Agent Benchmark, a product focused on evaluating how open-weight models perform on long-horizon, real-world legal tasks. Che

We’ve developed our own inference engine Runtime-Optimized Serving Engine (ROSE) to serve models ranging from embeddings to trillion-parameter LLMs. With CuTeDSL integrated into our inference engine,

We’ve partnered with @AMD, @Broadcom, @Intel, @Microsoft, and @NVIDIA, to release Multipath Reliable Connection (MRC), a new open networking protocol that helps large AI training clusters run faster a

What if you could extract text from any photo on your phone? We built LlamaParse Mobile, an @expo + @reactnative app for iOS & Android, powered by the LlamaParse TypeScript SDK 📱 Three steps, that’s i

Sample-based quantum diagonalization is a quantum algorithm technique that estimates the eigenvalues and eigenvectors of matrices using measurements from quantum hardware, rather than requiring full q

arXiv:2603.12799v2 Announce Type: replace Abstract: Achieving adversarial robustness in Vision-Language Models (VLMs) inevitably compromises accuracy on clean data, presenting a long-standing and chal

This Reddit discussion explores which human skills will remain valuable and relevant in a future where AI systems can perform most technical tasks, likely covering interpersonal abilities, creative th

Open hybrid cloud is becoming the control plane that determines how enterprise artificial intelligence actually runs in production. The conversation has moved quickly from experimentation to execution

What’s going on in semis is not sustainable per GS Everyone spending on AI is losing money except for semiconductor companies and the dynamic is “unprecedented and unsustainable” “Something has to cha

The AI era demands a fundamental shift in security, and that includes identity and access management (IAM). Traditional controls simply aren’t built for autonomous AI agents that interact with sensiti

arXiv:2605.02187v1 Announce Type: cross Abstract: Bring-Your-Own-Key (BYOK) agent architectures let users route LLM traffic through third-party relays, creating a critical integrity gap: a malicious r

arXiv:2510.08952v4 Announce Type: replace Abstract: Text-attributed graphs (TAGs) have become a key form of graph-structured data in modern data management and analytics, combining structural relation

arXiv:2605.03096v1 Announce Type: cross Abstract: In classification tasks, models may rely on confounding variables to achieve strong in-distribution performance, capturing spurious features that fail

arXiv:2605.02914v1 Announce Type: new Abstract: A guard model fine-tuned on entirely benign data can lose all safety alignment -- not through adversarial manipulation, but through standard domain spec

arXiv:2605.02915v1 Announce Type: new Abstract: Same-model self-verification, prompting a model to audit its own predicted answer, is a plausible confidence signal for selective prediction, but its pr

arXiv:2605.02463v1 Announce Type: cross Abstract: Multi-agent LLM systems are increasingly used to solve complex tasks through decomposition, debate, specialization, and ensemble reasoning. However, t

arXiv:2605.03314v1 Announce Type: new Abstract: In single-stream autoregressive interfaces, the same tokens both update the model state and constitute an irreversible public commitment. This coupling

arXiv:2605.03609v1 Announce Type: cross Abstract: Large language models often display heterogeneous moral preferences across settings. We study inference-time steering toward a desired ethical framewo

arXiv:2605.02920v1 Announce Type: cross Abstract: Standard transformer architectures learn fixed slow-weight representations during training and lack mechanisms for rapid adaptation within an episode.

Why are four of the top 10 trending vendors AI inference platforms? Companies are getting selective about which model handles which task, and routing tokens accordingly. Read more from @arakharazian a

Wiki-Builder, a tool for creating LLM-powered wikis, was featured on Hacker News' front page, announced by DAIR.AI. The post indicates upcoming updates and improvements to make wiki creation more acce

arXiv:2605.03304v1 Announce Type: new Abstract: The European Union's Carbon Border Adjustment Mechanism (CBAM) creates a complex challenge for the interconnected European electricity market. Tradition

arXiv:2603.27343v2 Announce Type: replace Abstract: Existing large language models (LLMs) evaluations use fixed-difficulty benchmarks that cannot adapt as models improve, and rarely isolate specific c

Anthropic shared their Wordle game result for puzzle #1,781, solved in 3 attempts with a final correct answer shown by five green squares (all letters in correct positions). The emoji grid represents

arXiv:2605.03596v1 Announce Type: cross Abstract: Workspace learning requires AI agents to identify, reason over, exploit, and update explicit and implicit dependencies among heterogeneous files in a

arXiv:2605.03475v1 Announce Type: new Abstract: Evaluating generative video models remains an open problem. Reference-based metrics such as Structural Similarity Index Measure (SSIM) and Peak Signal t

xAI and SpaceXAI have just made Colossus 1 available to Anthropic to support Claude. This means more than 220,000 NVIDIA GPUs in one of the world’s largest and fastest-built AI superclusters are now h

yep, we are about to live in a world where nobody trusts anything. just like Putin always wanted. congratulations! ps it will suck @GaryMarcus Or very soon no one will take any photo or video seriousl

Yesterday @ElevenLabs announced they'd crossed >$500M ARR. Here's my interview with @matiii from @sequoia AI Ascent a few weeks ago. Mati is a founder who really inspires me. He shares stories about E

Cursor 3.3 introduces a context usage breakdown feature that allows users to view detailed statistics about how their AI agent consumes context tokens. This diagnostic tool helps identify context-rela

This article explains how retrieval quality directly impacts large language model performance in retrieval-augmented generation (RAG) systems, emphasizing that even advanced LLMs cannot produce better

REALSTAGRAM_ZIMG is a LoRA (Low-Rank Adaptation) module designed for Z-Image Turbo that applies subtle realistic styling to generated images. The tool is compatible with character LoRAs, allowing user

arXiv:2605.02960v1 Announce Type: new Abstract: Production LLM workloads increasingly serve discriminative tasks, such as classification, recommendation, and verification, whose answers are read from

arXiv:2605.01222v1 Announce Type: new Abstract: Signal Temporal Logic (STL) offers verifiable task specifications and is crucial for safety-critical control. Yet STL planning remains challenging: exac

Databricks describes their approach to handling massive-scale observability and monitoring infrastructure that processes 10 trillion samples daily, moving beyond conventional monitoring solutions. The

Replit hosted a live broadcast retrospective celebrating a decade of development and community achievements through their 'Buildathon' initiative, showcasing projects, innovations, and milestones from

19歳でフランスの大学を中退してパリで世界初の完全自動化パスタロボット・レストランを作った創業者が米サンフランシスコで新会社「Eridale」を立ち上げ ━━ Eridaleの戦略 ━━ ・NYCで最初の5店舗展開 ・その後数千店舗へ展開予定 ・自社運営で自社ロボットを展開 ・店舗単位の最終利益50%超を狙う ・コスト削減分を高品質食材投資に ・既存のレストラン・テックは「ロボットを導入する」だけ

arXiv:2605.00856v1 Announce Type: cross Abstract: Accurate and continuous estimation of cognitive workload is fundamental to creating adaptive human-machine systems. However, designing architectures t