Knowledge catalogue
arXiv:2605.16665v1 Announce Type: new Abstract: Continental-scale knowledge of subsurface temperature is limited by the cost and sparsity of borehole measurements, but such information is essential fo
arXiv:2605.17279v1 Announce Type: cross Abstract: Code merging is a significant challenge, particularly in large-scale projects. Existing solutions, including program analysis and machine learning, sh
arXiv:2605.16764v1 Announce Type: new Abstract: Convolutional neural networks (CNNs) have been extensively and successfully applied to the task of synthetic aperture radar (SAR) image change detection
arXiv:2605.15315v1 Announce Type: new Abstract: LLM-powered coding agents spend the majority of their token budget reading repository files, yet much of the retrieved code is irrelevant to the task at
arXiv:2605.15422v1 Announce Type: new Abstract: Modern RL post-training methods such as GRPO and DAPO train on N response sequences of R tokens sampled from a shared prompt of P tokens, but standard F
Yohei Nakajima discusses the lack of elegant primitive implementations for stateful agents, decision traces, and context graphs, which are frequently discussed concepts in AI but rarely demonstrated i
Gemma-4-31B-it-Pearl supports 32K context, configurable thinking, function calling, and JSON mode. This is Together AI’s first Pearl-powered endpoint. Eventually, we plan to expand our Pearl powered p
New article: a visual tour of recent LLM architecture advances, from Gemma 4 to DeepSeek V4. I focus on long-context efficiency tweaks like KV sharing, per-layer embeddings, layer-wise attention budge
Personal financial decisions are full of assumptions, & the AI, as smart as it is, makes some of these without enough context. In addition to the skills, a system prompt that drives the AI to intervie
Gemma-4-31B-it-Pearl supports 256K context, configurable thinking, function calling, and JSON mode. This is Together AI’s first Pearl-powered endpoint. Eventually, we plan to expand our Pearl powered
Gemma-4-31B-it-Pearl supports text and image input, 256K context, configurable thinking, function calling, and JSON mode. This is Together AI’s first Pearl-powered endpoint. Eventually, we plan to exp
ICYMI: 1️⃣ LangSmith Engine 2️⃣ SmithDB 3️⃣ Managed Deep Agents 4️⃣ LangSmith Sandboxes: Now Generally Available 5️⃣ Context Hub 6️⃣ LangSmith LLM Gateway 7️⃣ Sandboxes, Prebuilt agents, + free model
Fireworks Training Platform continues to expand. Today GLM 5.1 LoRA RL is now live via Training API: SFT, DPO, and full RL on a 200K context window → custom loss functions or smart defaults. No usage
Granite Embedding Multilingual R2 is an open-source multilingual embedding model released by IBM under the Apache 2.0 license, featuring a 32K context window and sub-100M parameters. The model is opti
OpenAI implemented improvements to help ChatGPT better understand and respond appropriately to context in sensitive conversations, such as those involving mental health, abuse, or other delicate topic
arXiv:2508.14780v2 Announce Type: replace Abstract: Compression-based dissimilarities (CD) offer a flexible and domain-agnostic means of measuring similarity by identifying implicit information throug
Introducing Managed Deep Agents: the fastest way to ship a production deep agent. ✅ Harness ✅ Context ✅ Code execution All managed for you. Deploy with a single line of code. https://www.langchain.com
arXiv:2602.04284v2 Announce Type: replace Abstract: Managing agent context (e.g., thought and observation) during multi-turn agent-environment interactions is an emerging strategy to improve agent eff
arXiv:2605.09644v1 Announce Type: new Abstract: Visual Geometry Grounded Transformer (VGGT) advances 3D reconstruction via scalable Transformer architecture, but the quadratic complexity of global att
arXiv:2605.08405v1 Announce Type: new Abstract: How do LLMs learn in-context? Is it by pattern-matching recent tokens, or by inferring latent structure? We probe this question using a toy graph random
arXiv:2603.21901v2 Announce Type: replace Abstract: Video subtitle removal aims to distinguish text overlays from background content while preserving temporal coherence. Existing diffusion-based metho
Found my limit: ~8 simultaneous code generation/review tasks. Beyond that, context-switching turns from leverage into chaos. AI coding doesn’t remove the need for WIP limits. It might even make them m
arXiv:2605.08295v1 Announce Type: cross Abstract: While random demonstration labels barely hurt in-context learning (Min et al., 2022), we show that homogeneous labels--even semantically valid ones--c
arXiv:2605.08832v1 Announce Type: new Abstract: Neural surrogate models for computational fluid dynamics (CFD) are typically trained as forward operators that map explicit problem specifications, such
Rly looking forward to this Deep Agents workshop at @LangChain Interrupt Learning everything I can about Context Management and Deep Agents are the best examples of OSS best practices Harness to Fleet
we just shipped delta channels in langgraph 1.2. as agents run longer and use more context, full-state checkpointing doesn't scale, but delta channel snapshots do. this new algorithm is now powering m
arXiv:2602.09789v3 Announce Type: replace Abstract: Scaling up model parameters has long been a prevalent training paradigm driven by the assumption that larger models yield superior generation capabi
arXiv:2512.10371v2 Announce Type: replace Abstract: The rapid development of mobile GUI agents has stimulated growing research interest in long-horizon task automation. However, building agents for th
DeepSeek V4 Pro brings long-context reasoning and SOTA coding performance to Together AI serverless. The next layer is serving it efficiently: KV cache, prefix reuse, hybrid attention, batching, kerne
reply to Hinton’s reply to me, for additional context: Dear @geoffreyhinton, I literally never said that AI systems “JUST regurgitate”; that’s plainly false. I don’t believe it, and I didn’t say it. (
arXiv:2605.07042v1 Announce Type: new Abstract: Large Language Model (LLM) agents are deployed in complex environments -- such as massive codebases, enterprise databases, and conversational histories
To be clear, I meant 'idea' in the context of science and engineering, in particular AI, as in 'theory' or 'invention', like a new theory of cognition or an idea for a new ML approach. I thought that
Downloading now... 1M token context window with supposedly usable coding agent capability all on a 128GB Macbook Pro is 🤯 🚨 OPEN SOURCE AI IS LITERALLY UNSTOPPABLE 🚨 The legendary founder of Redis (An
arXiv:2605.04995v1 Announce Type: new Abstract: We compare in-context learning with fixed queries and agentic learning with adaptive queries for uniform approximation of task families. We consider two
arXiv:2509.14274v2 Announce Type: replace Abstract: Large Language Models (LLMs) have demonstrated significant promise in formal theorem proving. In this study, we investigate the ability of LLMs to d
people are really starting to use voice to interact with AI, especially when they have a lot of context to dump. GPT-Realtime-2 comes to the API today; it is a pretty big step forward. (we are working
arXiv:2605.05176v1 Announce Type: new Abstract: Pre-trained transformers are able to learn from examples provided as part of the prompt without any weight updates, a remarkable ability known as in-con
🚨Alphabet GOOGL trades at 133x free cash flow. For context: its pre-COVID multiple was ~20x. And free cash flow hasn't grown since 2021. GQG Partners — one of the world's top institutional investors —
arXiv:2605.02123v1 Announce Type: cross Abstract: The increasing use of token-based representations in language-driven applications has motivated wireless token communication, where tokens are treated
Finance Search supports questions that use live market data: prices, fundamentals, earnings, filings, and market context. Agents can run valuation lookups, earnings recaps, and market monitors without
Claude Code on desktop includes a preview feature that allows users to provide context to Claude by attaching DOM elements or drawing directly with a pencil tool. This functionality streamlines the pr
arXiv:2605.00827v1 Announce Type: cross Abstract: Large Language Model (LLM) agents increasingly interact with external systems through tool-calling protocols such as the Model Context Protocol (MCP).
This post discusses practical frameworks for working with AI models, covering how context serves as infrastructure, taste functions as configuration, verification enables autonomous systems, and deleg
arXiv:2601.21714v2 Announce Type: replace Abstract: The evolution of Large Language Model (LLM) agents towards System~2 reasoning, characterized by deliberative, high-precision problem-solving, requir
arXiv:2605.02558v1 Announce Type: new Abstract: Accurate badminton stroke prediction is crucial for fine-grained sports analysis and tactical decision support. However, existing methods struggle to mo
deepagents you can run with a 'virtual filesystem' lets do lots of great context engineering tricks, without requiring an actual sandbox environment! @hwchase17 Where there's struggle is all of these
arXiv:2605.00482v1 Announce Type: new Abstract: Mobile network operators must monitor thousands of heterogeneous network elements across the radio access network and the packet core, each exposing hig
arXiv:2605.00069v1 Announce Type: new Abstract: Elastic distances like dynamic time warping (DTW) are central to time series machine learning because they compare sequences under local temporal misali
We’re excited to be part of the @pinecone Nexus launch! For AI agents to actually work in the enterprise, they need context from the content organizations rely on every day. Box and Pinecone Nexus mak
the lock-in isn't the harness -- it's the context pipeline feeding it. whoever owns how repo state gets pulled, ranked, and compressed into the attention window owns the developer, regardless of model
Trooper v2.1 is a tool that provides automatic fallback functionality from cloud-based LLM services to local Ollama instances when cloud quota limits are exceeded, incorporating context compaction to
Carl Franzen / VentureBeat: xAI launches Grok 4.3, featuring “always-on reasoning”, 1M token context window, and low API pricing, and releases a voice cloning suite called Custom Voices — While Elon M
Frameworks are dead. @jerryjliu0 of LlamaIndex called it — and then rebuilt the company around what actually matters: context. New episode of Beyond the Pilot: https://www.youtube.com/watch?v=HbXvX-Kt
Join us Tue 5/5: #DeepSeek-V4's hybrid attention + sparse MoE reduces KV cache up to 90%, enabling 1M-token context. We'll cover why that makes it great for agentic workflows, what it took to serve at
arXiv:2604.26622v1 Announce Type: new Abstract: Autonomous LLM agents increasingly operate in long-horizon, interactive settings where success depends on reusing experience accumulated over extended h
arXiv:2505.14808v2 Announce Type: replace-cross Abstract: The transformer's remarkable ability to perform in-context learning (ICL) has sparked a wide range of studies designed to understand its stren
arXiv:2604.26837v1 Announce Type: new Abstract: Long-context LLM serving is bottlenecked by the cost of attending over ever-growing KV caches. Dynamic sparse attention promises relief by accessing onl
A million-token context window is not a strategy. 🛑 Our Head of DevRel, @RoieSchwabco , explains why dumping data is killing your RAG performance: 📍 One needle in a haystack? Easy. 📍 Multiple needles?
After fixing correctness issues, we turned to the next bottleneck: Prefill throughput and GPU memory pressure in long-context Coding Agent serving. To address this, we introduced LayerSplit, a layer-w
As models, contexts, and workloads grow, hidden assumptions in inference infrastructure can surface as output anomalies. Reliability requires more than throughput, latency, and availability. It also r