Knowledge catalogue
// In-context Retrieval at Million-token Scale // Great study providing better understanding of retrieval at million-token scale. They run the first systematic study of in-context retrieval at the sca
Most agentic retrieval demos assume clean, well-structured documents. Enterprise reality is often different, consisting of messy PDFs where critical information is buried across tables, figures, and c
We've created a comprehensive Retrieval Harness for modern agentic retrieval in 2026. The harness provides a persistent data pipeline that can connect to a data source, index and update a large knowle
Agentic retrieval is changing the way retrieval-augmented applications are built, especially in domains like legal and fintech, where agents need to autonomously navigate large, evolving knowledge bas
This video provides a quick explanation of Retrieval-Augmented Generation Models (RLMs), outlining their benefits and demonstrating their implementation with DeepAgents. The content likely covers how
so we built psql_bm25s. exact BM25 retrieval. native Postgres access method. ~23x faster than pg_search on the standard benchmark. retrieval stops being a budget item. the harness stops rationing. the
我們開源了這顆星球🌎上速度最快的低成本 bm25 引擎。 so we built psql_bm25s. exact BM25 retrieval. native Postgres access method. ~23x faster than pg_search on the standard benchmark. retrieval stops being a budget item. the
// Skill Retrieval Augmentation for Agentic AI // Great read for AI devs. (bookmark it) It's on finding efficient ways to incorporate skills for agents. The work introduces Skill Retrieval Augmentatio
This is a great tutorial (credits @itsclelia + @lancedb) on how to build a practical retrieval pipeline that integrates directly with your agent harness. 1. Ingest a massive pile of docs with litepars
Impressive paper showing how much the first retrieval step matters for deep research agents. It helps to improve GPT-5.5 from 83.1% to 90.5% on BrowseComp-Plus with the same retriever and the same age
We're introducing Q2D-Web (Query2Doc-Web), a benchmark and public leaderboard for evaluating retrieval in agentic RAG systems. Q2D-Web tests how embedding models perform on large-scale web search usin
We're excited to introduce the Retrieval Harness in LlamaParse - which is the 2026 version of RAG over documents Generalized agents need the right set of tools to scalably search and read through an a
Semantic search alone doesn't cut it. Neither does brute-force grep. Agents need both. Today we're shipping the Retrieval Harness in LlamaParse Index: semantic search, server-side grep, and file-level
Finally, a good paper testing if graph memory actually beats flat retrieval for long-term agents. (bookmark this one) Researchers extract each conversational turn into typed nodes and attributed edges
longmemeval experiment arch: 1) deterministic ingestion/extraction (85.6% accuracy, 86.2% retrieval) 2) semantic ingestion/extraction (84.8% accuracy, 94.9% retention) 3) semantic ingestion/determinis
Best part are retrieval and search details: >Agent queries differ from human queries & what good results means changes too >Parallel queries and ranking are both tools to the same outcome >Top-K preci
memory isn't a retrieval widget. it's write policy. the harness decides what survives compaction, what gets promoted, and what becomes reusable state. https://x.com/hwchase17/status/204297850056760973
I was unable to retrieve the specific content of the tweet at the URL provided (https://x.com/GaryMarcus/status/2042756801201082637). X (formerly Twitter) content is generally not accessible via we...
Recommended if you work on agent memory. Retrieval is the hardest memory problem in most harnesses because agents keep pulling stale context. AML tests this with a coding track of 150 software tasks,
Memory operations, including retrieval, prioritization, compaction awareness, should be native and baked into the harness. 𝐖𝐢𝐭𝐡𝐨𝐮𝐭 𝐭𝐡𝐞 𝐩𝐫𝐨𝐩𝐞𝐫 𝐢𝐧𝐭𝐞𝐫𝐚𝐜𝐭𝐢𝐨𝐧 𝐛𝐞𝐭𝐰𝐞𝐞𝐧 𝐡𝐚𝐫𝐧𝐞𝐬𝐬 𝐚𝐧𝐝 𝐦𝐞𝐦𝐨𝐫𝐲, 𝐦𝐞𝐦𝐨𝐫𝐲 𝐚𝐥𝐨𝐧𝐞 𝐢𝐬 𝐩𝐨
retrieval is a decision problem, not just a similarity problem cool experiment from @assaf_elovic swapping embeddings for Jev in GPT Researcher - 73% vs 46% relevant context decision models will show
// Neural procedural memory // Good paper on agent memory beyond prompt retrieval. NPM stores procedural skills as activation steering vectors distilled from contrastive historical experience. Textual
Our Founder and Chief Scientist @EdoLiberty just kicked off the Search & Retrieval track @aiDotEngineer. 'Agents don't need to be smarter. They need a Knowledge Layer.' We've got a big update on this
agree a lot with this - the actual storage layer for memory (eg which database you use) matters less than the process around that data (retrieval, ingestion, etc) @hwchase17 what if memory is not stor
Four weeks, three enterprise customers, same pattern across all three. Most inference spend goes to retrieval loops. A generic index carries no knowledge of domain, query types, or task structure, so
Great to see a GBrain x ActiveGraph crossover here Great retrieval pairs well with almost anything you want to do with agents Gbrain is what an agent *knows* — a durable markdown/git knowledge substra
this is a fantastic example of a robust agentic memory system built on langgraph! there's 4 key parts 1. retrieval (rag) 2. storage, fragmented by memory type 3. reasoning* 4. learning (often called d
Real question: what is the actual latest state-of-the-art for file search and retrieval? - Actual grep over filesystem - Virtualized grep / BM25 over a db (what @mintlify did) - Vector search over a d
Releasing my first kernel on @huggingface: MaxSim Late-interaction retrieval (ColBERT / PyLate) bottlenecks on materializing the full similarity matrix. This kernel avoids it by using tiled scoring wi
A beginner-friendly guide to building a simple RAG (RETRIEVAL-AUGMENTED GENERATION) AI Agent with Pinecone + n8n 🧵 If you’re learning AI automation and you’re confused about Pinecone, knowledge bases,
A preview of Pinecone Nexus, a data retrieval engine available through an early access program, uses the declarative KnowQL language to improve accuracy while reducing token consumption by up to 90%.
Its noticeable how much of the whole practice of working with AI - the prompts, the skill files, the connectors, retrieval work, the markdown files, etc. - is a substitute for the real problem of cont
Harrison Chase, co-founder of LangChain, argues that while prompt engineering is foundational and can get you 80% of the way to a functional AI system, the remaining 20% of engineering effort involves
Memory is where the harness stops being a wrapper and becomes an ownership layer. Once it controls what gets remembered, retrieved, compressed, and acted on, it starts shaping the agent’s judgment, no
// When to Retrieve During Reasoning // Pay attention to this one, AI devs. (bookmark it) Most RAG systems retrieve once, before the model starts reasoning. Large reasoning models like o1 and R1 don't
Good measurement work on whether retrieved agent skills actually help. They report that agent skills that lift your aggregate score can be hurting every task they touch. The usual way of checking comp
Jörg Schad is the VP of Engineering at @pinecone. He joins @kbal11 to discuss moving beyond RAG with precomputed context, context artifacts, semantic layers, and AI retrieval. @joerg_schad https://sof
Nice paper combining the strength of Skills and RAG. Most RAG systems retrieve on every query, whether the model needs help or not. This is wasteful when the model already knows the answer, and often
Most AI agents reset every session. @JenovaAIAgent's don't. Longest session on their platform: 16M tokens. All of it retrievable in <10ms via Pinecone vector retrieval. Result: Fast ramp to $1M+ ARR,
Yann LeCun, Executive Chairman of AMI Labs, explains why LLMs are mostly retrieving human knowledge rather than thinking for themselves: LeCun starts with why so many people misread what these systems
Fine-tune your own embedding model for the price of a coffee. A great reranker can't surface a doc that was never retrieved. A RAG pipeline cannot cite a case it failed to retrieve. See how contrastiv
Very cool idea to convert memory to skills. (bookmark it) Most agent memory systems retrieve past traces as passive context. MSCE turns them into executable skills instead. The training-free framework
Building search across text, images, audio, and video? Multimodal search uses vector embeddings to retrieve content based on meaning, not just keywords or metadata. Less manual tagging. More flexible
We're releasing pplx-embed-v2-late, two late-interaction embedding models that retrieve text, images, and pages with a shared embedding space for cross-model querying. Both models achieve frontier per
We're sharing new research on how models hack public benchmarks. The latest models, including Opus 4.8 and Composer 2.5, learn to retrieve solutions from the internet or git history. When we apply a s
// Memory as a Model // The paper augments any LLM with a separate trained memory model that stores, retrieves, and integrates facts on its behalf. It decouples memory updates from base-model weight u
Finance Search is now available in the Perplexity Agent API. In one tool call, developers can now retrieve licensed financial datasets, real-time market data, and cited web sources for agents that nee
Simply retrieving a reasoning trace looks a lot like human reasoning, until it's time to navigate uncharted territory. If you memorized all reasoning traces of humans from 10,000 BC, you could automat
// Multi-Agent Synthesis RAG // Nice paper on improving RAG systems with multiple agents. (bookmark it) The paper introduces MASS-RAG, a multi-agent synthesis framework for retrieval-augmented generat
Nice paper on agent memory. It explores an effective routing mechanism for memory in agents. The overall finding is that you want a separate retrieval strategy for each type of memory instead of one e
Obligatory plug for static embeddings. Are you using static embeddings in your projects? Would love to hear how people are leveraging them, whether for retrieval and beyond. Static embeddings offer un
Static embeddings offer unmatched throughput, but they also suffer from accuracy loss compared to traditional embedding models. Can we make them better for retrieval? We tried several things: ✅️ raw m
Voyage AI models now run natively on Fireworks, the first and only dedicated inference platform @VoyageAI by @MongoDB has partnered with. Embed, retrieve, rerank, generate: your full retrieval pipelin
I wasn't able to retrieve the content of that specific tweet or URL from the search results. X (Twitter) posts are generally not indexed in a way that allows direct retrieval of their content, and ...
I was unable to retrieve the specific tweet at the URL provided (status ID 2042072653658460260). This appears to be a future-dated or non-existent post — the status ID is beyond what currently exis...
I was unable to retrieve the specific post at that URL. The status ID `2041761154419380444` does not appear in any search results, and X (formerly Twitter) requires JavaScript to render individual ...
This is all I will say on the matter: It’s alarming to see the constant effort to try to tell me how I should use AI, how I should build AI, and now how I should pace it. Is it just me, or are all the
A research agent is only as good as the search layer behind it! I built a Deep Research Agent using LangChain Deep Agents with a search-first architecture designed around grounded, source-backed resea
Does your context compactor actually save anything? If you built or tuned one, this one is worth your time. (bookmark it) Task completion is the standard way to validate compression, and it hides the
Finally a good paper testing whether agent memory needs an LLM at all. Production memory stacks spend extra model calls on summarizing interactions, writing records, and reranking retrievals. Every on