News
· 2026-11-16
Novo Nordisk's Manufacturing Technology team has deployed a process-intelligence assistant on AWS for a vault containing roughly four million controlled documents. Amazon Bedrock AgentCore runs the query agent; Amazon OpenSearch Serverless combines BM25 with Titan dense embeddings, while Amazon Neptune stores the process hierarchy, manufacturing entities, document lifecycle state, and cross-document relationships. Amazon Bedrock Rerank then refines the fused candidates before the agent answers w
News
· 2026-11-16
Amazon Bedrock Managed Knowledge Base now adds tabular search to agentic retrieval. Rather than treating a spreadsheet as a set of passage-sized chunks, the service first retrieves the relevant file, lets the agent generate code, and executes that code over the original CSV, XLSX, or XLS file inside an isolated sandbox without network access. Only printed results return to the agent, which combines them with retrieved policy or memo passages and emits a cited answer.
This design restores whole-
News
· 2026-11-16
Amazon Web Services has added managed retrieval instructions to Amazon Bedrock Managed Knowledge Base, allowing each tenant to define a persona, answer format, retrieval guidance, and answer policy while sharing retrieval infrastructure. Instructions are asynchronously reviewed, versioned, and applied through `AgenticRetrieveStream`; Planning trace events expose the exact instruction version used.
The control plane does not replace tenant isolation. Bedrock applies the filter supplied in a requ
· 2026-11-14
SparseRead is a reusable control layer that helps AI agents retrieve only the external evidence needed for a task, without modifying the underlying agent model. Its Read Intent Protocol tracks evolving requirements and source coverage, Adaptive Reader Backends choose how evidence is located, and a Budgeted Read Gate caps cumulative exposure while retaining a faithful full-view fallback.
Across six models and five scenarios, the paper reports up to 92.9% fewer total tokens and 89.0% lower wall t
News
· 2026-11-13
GPBam evaluates legal retrieval and long-form legal writing with 81 private German public-law examination cases, reference Gutachten solutions, and a separate recitation task covering more than 200 statutory provisions. Without retrieval, 32 open-weight and proprietary models span 4.81 to 76.42 out of 100. Adding retrieved law produces nearly threefold variation in retrieval performance, yet improves essay scores by at most 4.1 points and yields no significant gain for three of seven tested mode
News
· 2026-11-12
DeepStress is a dynamic stress-testing framework for deep search agents. It replaces the retrieval stack with a synthetic environment that generates documents for each query while independently controlling their trustworthiness, relevance, and factuality. This makes evidence degradation reproducible instead of relying on rare failures buried in static benchmark corpora.
The framework introduces a five-way outcome taxonomy and two reliability-aware metrics, RAS and CoPRAS, that reward cautious b
News
· 2026-11-12
Natera and AWS have deployed SWARM, an agentic AI system for clinical laboratory operations, on Amazon Bedrock AgentCore. Its architecture separates two execution modes: swarm-agent performs autonomous, high-throughput diagnosis for incoming Jira tickets, while swarm-hive supports engineers with interactive troubleshooting, image uploads, and persistent conversation context.
The system combines a versioned library of 33 deterministic and LLM-powered skills, an equipment-aware RAG pipeline, even
News
· 2026-11-12
LlamaIndex has released ParseBench-2, a document-parsing benchmark designed to evaluate semantic damage rather than treating every representation difference as an equal failure. The corpus expands from 2,100 to 4,300 cases and covers text, tables, formatting, charts, layout, and forms, including complete validation of all 1,025 charts.
At the center of the benchmark is a catalog of more than 330 injected error scenarios, each classified as critical, major, minor, or harmless. These severity lab
News
· 2026-11-11
AWS has consolidated 23 production-oriented generative AI building blocks from the AWS Generative AI Innovation Center into an open-source catalog spanning RAG, document processing, contact-center agents, agent scaffolding, evaluation, and operations. Each asset includes deployment documentation and infrastructure-as-code templates, with prior customer deployments providing evidence of delivery gains: DoorDash reportedly expanded testing capacity 50-fold, while Ricoh increased document-processin
News
· 2026-11-11
Hugging Face introduces token-pooling-aware training for multi-vector embedding models. By training models with the same aggregation behavior expected at inference time, the approach aims to preserve retrieval quality while reducing the computational and storage burden associated with representing every token as a separate vector. The method is particularly relevant to retrieval-augmented generation and semantic search systems where multi-vector representations improve recall but can make indexi
News
· 2026-11-10
LangChain outlines retrieval and context-engineering techniques used by its Deep Agents harness to address inefficient search, oversized tool results, and stale data. The central idea is to treat retrieval as part of agent orchestration rather than as an isolated vector-search call. Results should be selected, compressed, deduplicated, freshness-checked, and budgeted according to the active task. In production, retrieval quality must be measured through end-to-end task success, groundedness, lat
News
· 2026-11-09
HiVe-Mem introduces a two-dimensional memory framework for multimodal agents operating over long interaction histories. Its horizontal expansion builds a relational graph over episodes using textual and visual anchors, allowing semantic retrieval seeds to recover complementary information from neighboring memories. Its vertical evidence composition retains compact summaries, captions, grounded visual views, raw images, and interaction traces, then uses a lightweight trainable selector to assembl
News
· 2026-11-07
SparseRead is a reusable control layer that regulates what external content enters an AI agent’s context without modifying the underlying model. Its Read Intent Protocol tracks evolving evidence needs and source coverage, Adaptive Reader Backends accommodate heterogeneous resources, and a Budgeted Read Gate limits cumulative raw exposure while preserving a faithful full-view fallback.
Across six models, five scenarios, and three agent frameworks, the system reportedly reduces token usage by as
News
· 2026-11-06
Researchers introduce KG-RAG4SM, a knowledge-graph retrieval-augmented framework for schema matching. It combines LLM-based, vector-based, graph-traversal, and hybrid retrieval with reranking to extract relevant subgraphs from external knowledge graphs, without retraining the underlying language model. On MIMIC, the method improves precision by 39.79% and F1 by 30.56% over Jellyfish-8B; on Synthea, it improves precision by 69.20% and F1 by 21.99% over SMAT.
A particularly important design choic
News
· 2026-11-06
RAGCompass addresses the fact that no retriever is consistently best for every query in retrieval-augmented generation. Rather than routing from query features alone, it compares the passages produced by each retriever and fine-tunes a reranker against the downstream answer each evidence set yields. A small comparison head provides an additional, moderate gain, while neither component contains retriever-specific parameters.
The paper reports a 28% relative improvement in answer F1 over the best
News
· 2026-11-06
The DS@GT ARC team developed a four-stage fact-checking article generation pipeline for CLEF 2026 CheckThat! Lab Task 3. It cleans noisy web evidence, retrieves and clusters relevant sources, generates an article with Llama-3.1-8B-Instruct conditioned on the original fact-checker verdict, and restores inline citations through deterministic post-processing.
The system achieved a mean score of 0.428 and ranked third among nine teams, compared with a 0.272 baseline. Its main engineering contributi
News
· 2026-11-05
The paper introduces a lightweight optimization for retrieval embeddings: prepend trainable [DUMP] tokens to every input and update them through the standard contrastive retrieval objective. The model architecture, pooling strategy, and loss remain unchanged. By interacting with regular tokens during encoding, these prefix positions appear to absorb or redirect residual token-level signals associated with attention sinks and low-information tokens, producing cleaner fixed-dimensional retrieval r
News
· 2026-11-05
AWS has announced the general availability of Web Answer on Amazon Bedrock AgentCore. This managed connector accepts a natural-language question, plans and executes parallel searches, reads key pages to verify facts, and searches again when the evidence is insufficient. It then returns a synthesized answer whose claims are linked to source URLs through a single Model Context Protocol tool call.
The service packages search, page retrieval, synthesis, and citation attribution into an AWS-operated
News
· 2026-11-05
Apple Machine Learning Research has published work on accelerating maximum inner product search (MIPS), a core retrieval primitive used by recommendation systems, search engines, and retrieval-augmented generation (RAG). The central challenge is that exhaustive comparison becomes prohibitively expensive when the index contains millions of vectors with thousands of dimensions, while conventional tree-based search is difficult to apply in high-dimensional spaces.
The research points to a need to
News
· 2026-11-05
RAGCompass addresses a central limitation of multi-retriever RAG systems: no single retriever is consistently best for every query. Rather than routing from query features and a fixed retriever inventory, the method directly compares the passages returned by each retriever. It contrastively fine-tunes a reranker against the answer produced from each evidence set, then adds a lightweight comparison head to improve candidate ordering without retriever-specific parameters.
The paper reports a 28%
News
· 2026-11-05
A medical multi-agent study replaces question-time retrieval with adaptive intervention driven by gaps observed during reasoning. An external CHALLENGER inspects the evolving trace and triggers targeted knowledge retrieval or specialist recruitment only when a factual or expertise gap is detected.
On 862 hard questions from nine biomedical datasets, in-between retrieval improved accuracy on eight of nine datasets and exceeded Challenger-only performance by six percentage points. Challenger-trig
News
· 2026-11-05
DS@GT ARC has developed a four-stage pipeline for CLEF 2026 CheckThat! Lab Task 3. It cleans noisy scraped evidence, retrieves and clusters relevant sources, conditions Llama-3.1-8B-Instruct on the original fact-checker verdict to generate the article, and restores inline citations through deterministic post-processing. The system ranked third among nine teams, achieving a mean score of 0.428 versus a 0.272 baseline. Its central design choice is to separate verdict determination from article gen
News
· 2026-11-04
Researchers apply FlyHash—an LSH-inspired scheme derived from fruit-fly olfactory circuits—to retrieval for LLM agent memory. Across 10 random seeds, four datasets, and two modalities, FlyHash preserves neighborhood structure more efficiently per bit than SimHash and remains superior when evaluated against genuine, non-circular relevance labels rather than embedding-derived nearest neighbors.
The engineering result is narrower than the headline quality result. At equal bit budgets, FlyHash prod
News
· 2026-11-04
Indegene has built a production-oriented Data Quality Agent within its NEXT Unified Data Platform on AWS. The system uses Claude Sonnet 4.5 on Amazon Bedrock, Bedrock AgentCore, Strands Agents, and two RAG knowledge bases to combine table metadata, profiling results, anomaly signals, existing rules, external standards, and internal enterprise definitions. It can propose column-level, table-level, and natural-language-to-SQL validation rules before generating executable SQL and Python.
The archi
News
· 2026-11-03
HERIT introduces a domain-adaptation pipeline for translating historical Hanja documents into English. It uses retrieval-augmented generation to create pseudo-labeled training data from abundant Hanja-Korean corpora, expanding supervision while reducing the temporal bias caused by scarce, period-skewed human translations. The paper reports consistent gains over baselines in both automated and human evaluations and uses the resulting system to translate previously inaccessible archival material.
News
· 2026-11-03
arXiv has released a comprehensive survey on Text-to-SQL systems, focusing on the latest advancements driven by large language models (LLMs). The report systematically reviews the evolution from traditional rule-based methods to modern LLM architectures, highlighting systems like ReFoRCE, Agentar-Scale-SQL, Arctic-Text2SQL-R1, and XiYan-SQL. These systems leverage modular design, multi-agent collaboration, and retrieval-augmented generation to significantly improve the accuracy of complex query
News
· 2026-11-03
RandHan has released ResMem, a novel model that transforms retrieved knowledge into parametric memory through offline teacher guidance, reducing online retrieval costs and enhancing model performance. By leveraging a large language model (LLM) as an offline teacher to generate teacher targets, ResMem trains a parametric memory to learn an additive logit residual. This approach significantly improves mean token F1 scores across five question-answering datasets while reducing the breakage of the L
News
· 2026-11-02
TRIO decomposes retrieval-augmented generation into three cooperating agents for query expansion, retrieval-sufficiency assessment, and answer generation. Each agent is optimized around a distinct but aligned objective: answer quality rewards query expansion, unanimity voting and generated outcomes supervise the judge, and dynamic matching rewards combined with curriculum learning improve evidence use in the final answer.
The design targets three recurring RAG failure modes: ambiguous queries,
News
· 2026-11-02
Hugging Face has released a Retrieval-Augmented Generation (RAG) system tailored for stock market information retrieval. By integrating retrieval-based and generative approaches, the system enhances the accuracy and relevance of financial text generation. Evaluated using the ROUGE metric on a dataset of 100 stock market-related queries, the system demonstrates high ROUGE scores and an average response time of 1.16 seconds, highlighting its potential for efficient and accurate financial informati
News
· 2026-11-02
A study published on arXiv investigates the performance of Retrieval-Augmented Generation (RAG) models when exposed to a majority of mutually agreeing poisoned documents. The research demonstrates that the detection rate of misinformation by the Gemma-3-12b model decreases significantly from 25% in the minority poisoned condition to 10% in the majority condition. The experiment, validated by the McNemar test (p = 6.04e-7), reveals a conformity-like bias in the model's response to a majority of p