Retrieval-Augmented Generation
retrieval-augmented-generation-92a54a4a·23 events·first seen 1mo agoAliases: Retrieval-Augmented Generation, Retrieval-Augmented Generation (RAG)
Co-occurring entities
More like this (12)
Guides (1)
Recent events (23)
Separating Semantic Competition from Context Length in RAG Reading
This paper introduces a matched-control protocol to isolate whether RAG reader failures stem from context length or semantic competition among retrieved passages. By replacing hard-competitor passages with less competitive ones while holding passage count and length fixed, the authors demonstrate a measurable competition effect on SQuAD using Phi-2 and Qwen2.5-1.5B. Phi-2 recovers +6.0 EM and +7.0 answer-inclusion points; Qwen2.5-1.5B recovers +4.5 EM and +9.0 answer-inclusion points. The study also introduces retention curves and a right-censored half-life metric to track performance degradation as competitors accumulate.
LEANN: RAG System with 97% Storage Savings for On-Device Private Retrieval
LEANN is an open-source retrieval-augmented generation (RAG) system targeting personal device deployment with claimed 97% storage reduction compared to conventional vector index approaches. The project is associated with MLsys 2026, suggesting an upcoming systems research paper. It emphasizes privacy through fully local execution and aims to maintain retrieval accuracy despite aggressive compression. The repository has accumulated over 11,000 stars with strong recent momentum.
Expert Support Case Study: Bolstering a RAG App with LLM-as-a-Judge
Hugging Face published a case study describing how Digital Green used an LLM-as-a-Judge approach to evaluate and improve a retrieval-augmented generation (RAG) application. The post covers the methodology for using LLMs to score and validate RAG outputs, providing a practical deployment pattern for quality assurance in production AI systems. It serves as a concrete example of enterprise-grade evaluation pipelines built on top of RAG architectures.
Building Cost-Efficient Enterprise RAG Applications with Intel Gaudi 2 and Intel Xeon
This Hugging Face blog post details how to build retrieval-augmented generation (RAG) pipelines for enterprise use cases using Intel Gaudi 2 accelerators and Intel Xeon CPUs. It covers the architecture and cost-efficiency tradeoffs of deploying RAG on Intel hardware as an alternative to GPU-based infrastructure. The post is positioned as a practical guide for organizations seeking lower-cost inference deployments.
Retrieval-Augmented Detection of Abusive Clauses in Chilean Terms of Service
Researchers present a RAG framework for automated detection and classification of potentially abusive clauses in Chilean Terms of Service agreements, designed for local execution with open-weight language models. They introduce the Chilean Abusive Terms of Service Extended corpus with 100 contracts and 10,029 annotated clauses across 24 legally grounded categories. Experiments show RAG prompting substantially improves performance, enabling local models to approach larger cloud-based systems at reduced computational and token cost. The work also contributes a refined legal annotation scheme for AI-assisted consumer contract review.
Factual Density (FD*): A Retrieval Optimization Signal for Multi-Source RAG in Medical AI
This paper introduces Factual Density (FD*), a retrieval reranking signal that measures the proportion of verified atomic claims per token to address what the authors call the 'Expert Blindness Effect' in standard RAG pipelines. Using the NexusAgentics Ghost Audit preprocessing pipeline and Z-score normalization within length bins, FD* is validated as a length-independent signal. Evaluated on the HealthFC benchmark (750 health claims), FD*-optimized retrieval achieved 100% systematic review saturation in top-5 results, surfacing Cochrane evidence that cosine similarity ranked outside the top ten. The study is limited to 25 verified mappings across seven claims, with full n=50 validation deferred to future work.
awesome-llm-apps: 100+ Runnable AI Agent & RAG Application Examples
A curated GitHub repository collecting over 100 deployable AI agent and RAG (Retrieval-Augmented Generation) applications built with LLMs. The collection is designed for practical use — clone, customize, and ship. With 110,915 total stars and 202 added today, it reflects strong community interest in applied LLM tooling.
Generative AI Advertising as a Problem of Trustworthy Commercial Intervention
This paper argues that generative AI fundamentally transforms advertising by enabling interventions on the generative process itself rather than discrete content placement. The authors introduce a taxonomy of influence tiers—product mentions, information framing, behavioral redirection, and long-term preference shaping—and analyze how these manifest across RAG and agentic pipelines. They find that deployed systems focus on the most observable tier while more consequential, latent forms of commercial influence lack detection, measurement, or disclosure frameworks. The central challenge posed is whether commercial influence in generative systems can be made attributable, measurable, contestable, and aligned with user welfare.
BRANE: Natural Language Query-to-Configuration Selection for Retrieval Agents
BRANE is a system that dynamically selects retrieval agent pipeline configurations (LLM, retriever, number of hops, synthesis strategy) at inference time based on per-query characteristics and a cost-quality target. It uses an LLM to extract workload features from each query, then applies lightweight per-configuration predictors to estimate correctness, selecting the configuration that maximizes predicted accuracy penalized by cost. Evaluated on MuSiQue, BrowseComp-Plus, and FinanceBench, BRANE matches best-fixed-configuration accuracy at up to 89% lower cost and outperforms LLM-routing and fine-tuned Qwen3-4B baselines. The work frames per-query pipeline configuration as a practical alternative to static workload-level tuning.
Signal Collapse and Reward Hacking in Checker-Guided RAG for Biomedical QA
This paper investigates why NLI-based claim checkers used as process rewards in RL-trained medical RAG agents succeed or fail during training. The authors find that a checker's output distribution during training—not its held-out accuracy—determines whether it provides useful gradient signal, with LLM log-probability scoring causing near-total signal collapse (97%+ neutral labels) while a calibrated MedNLI classifier avoids this. A key finding is that stronger checkers can trigger reward hacking cascades (ultra-short answers, search avoidance, language collapse), while moderate-signal local classifiers yield better final model quality (+12% BERTScore over zero-shot). The work frames these as boundary conditions for verifier-as-reward systems in RLVR pipelines.
Production Agentic RAG Course Repository Gains Traction on GitHub
A GitHub repository titled 'production-agentic-rag-course' by jamwithai has accumulated 6,158 stars with 45 added today, indicating community interest in production-grade agentic retrieval-augmented generation systems. The repository appears to be an educational resource focused on deploying agentic RAG pipelines in production environments. Its trending status reflects ongoing developer demand for practical guidance on agentic and RAG architectures.
RASER: Recoverability-Aware Selective Escalation Router for Multi-Hop Question Answering
RASER introduces a family of lightweight routers that decide whether to escalate retrieval complexity for multi-hop QA without making additional LLM calls. Built on top of one-shot RAG using six derived features, RASER-2 and RASER-3 route queries to progressively more expensive retrieval strategies (PRUNE, IRCoT) only when needed. Across six LLMs and three benchmarks, the routers match SOTA F1 while consuming only 41-49% of the tokens required by always-escalating baselines.
LongMINT: Benchmark for Evaluating Memory Under Multi-Target Interference in Long-Horizon Agent Systems
LongMINT is a new benchmark designed to evaluate memory-augmented agents in realistic long-horizon settings where information is repeatedly updated and interferes across memories. It contains 15.6k QA pairs over contexts averaging 138.8k tokens (up to 1.8M tokens), spanning domains including state tracking, multi-turn dialogue, Wikipedia revisions, and GitHub commits. Evaluation of 7 representative systems—including vanilla long-context LLMs, RAG, and memory-augmented agent frameworks—reveals consistently low average accuracy of 27.9%, with performance particularly degraded on multi-target aggregation tasks and when earlier facts are revised by subsequent context. The analysis identifies retrieval and memory construction as the primary bottlenecks.
Blue J Uses GPT-4.1 and RAG to Scale Tax Research for Legal Professionals
Blue J has built AI-powered tax research tools on top of OpenAI's GPT-4.1, combining domain expertise with Retrieval-Augmented Generation to deliver cited tax answers. The platform serves tax professionals across the US, Canada, and the UK. This is a case study published by OpenAI highlighting enterprise deployment of GPT-4.1 in a regulated professional domain.
WebGPT: Improving the factual accuracy of language models through web browsing
OpenAI fine-tuned GPT-3 to answer open-ended questions more accurately by giving it access to a text-based web browser. The system, called WebGPT, uses reinforcement learning from human feedback to learn to search the web, read pages, and cite sources. This work represents an early demonstration of retrieval-augmented generation and tool-use in large language models.
Maat: ReAct-Based Agentic Legal Research Assistant for Competition Law
Maat is a ReAct agent designed specifically for competition law research, orchestrating tools for RAG-based retrieval, web search fallback, and citation generation. Built iteratively with domain experts, it addresses hallucination and citation gaps found in general assistants (Claude, ChatGPT) and legal-specific models (SaulLM-7B, LegalGPT). Maat significantly outperforms baselines on case-specific tasks and matches top baselines on theoretical questions. The evaluation dataset is publicly released on GitHub.
Systematic Study of Schwartz Value Detection in Political Texts: Context, Scale, and Moral Knowledge
This paper investigates when additional context, larger models, or retrieved moral knowledge improve detection of Schwartz human values in political text using the ValueEval benchmark format. Key findings show that full-document context helps supervised DeBERTa encoders (+3.8–4.8 macro-F1) but not zero-shot LLMs, while RAG with a curated moral knowledge base consistently benefits all model families under early fusion. Scaling model size does not guarantee gains, and simple early fusion outperforms more complex RAG variants. The study recommends jointly evaluating context, knowledge, and model family rather than assuming larger inputs or models universally improve value-sensitive NLP.
MemTrace: Framework for Tracing and Attributing Errors in LLM Memory Systems
MemTrace introduces a framework that converts LLM memory pipelines into executable memory evolution graphs to enable fine-grained error tracing and root-cause attribution. The authors construct MemTraceBench, a benchmark covering Long-Context, RAG, Mem0, and EverMemOS memory systems, to systematically characterize memory failure modes such as information loss and retrieval misalignment. An automatic attribution method iteratively traces operation subgraphs to pinpoint failures, and the resulting signals are used to guide prompt optimization in a closed-loop system that improves end-task performance by up to 7.62%.
Training and Finetuning Reranker Models with Sentence Transformers
Hugging Face published a tutorial on training and fine-tuning reranker models using the Sentence Transformers library. Rerankers are cross-encoder models used in retrieval-augmented generation (RAG) and search pipelines to re-score candidate documents for improved relevance. The post covers dataset preparation, loss functions, and training configurations specific to reranking tasks.
Deploy Embedding Models with Hugging Face Inference Endpoints
Hugging Face published a guide on deploying embedding models using their Inference Endpoints service. The post covers how to set up dedicated endpoints for embedding models, enabling scalable vector generation for downstream tasks like semantic search and retrieval-augmented generation. This is part of Hugging Face's broader push to make production deployment of specialized model types more accessible.
DeepWeb-Bench: A Hard Deep Research Benchmark Requiring Cross-Source Evidence and Long-Horizon Derivation
DeepWeb-Bench is a new benchmark designed to stress-test frontier language models on deep research tasks—open-web search, evidence collection, and multi-step derivation—where existing benchmarks have become saturated. The benchmark evaluates nine frontier models across four capability families (Retrieval, Derivation, Reasoning, Calibration) and finds that retrieval is not the primary bottleneck; derivation and calibration failures account for over 70% of errors. Strong models fail via incomplete derivation while weak models fail via hallucinated precision, and models show genuine domain specialization with low cross-model agreement (rho = 0.61). The benchmark, rubrics, and evaluation code are publicly released.
DeepLearning.AI Launches AI Andrew: A Personality-Shaped AI Companion Built on Agentic Harness
Andrew Ng's team at DeepLearning.AI has released 'AI Andrew,' an AI companion designed to emulate Ng's communication style and personality for conversations about AI, careers, and learning. The system uses an agentic harness combining RAG, small and large models, guardrails, short- and long-term memory, and offline agentic loops that automatically propose system improvements. The team employed iterative error analysis to close the gap between AI Andrew's outputs and Ng's actual communication style, though acknowledged remaining issues including hallucinations. The product targets people seeking guidance on AI concepts, career decisions, and project ideas.
SECDA-DSE: LLM-guided design space exploration for FPGA accelerator generation
SECDA-DSE is a framework that integrates LLMs into the SECDA hardware-software co-design ecosystem to automate design space exploration (DSE) of FPGA-based AI accelerators. The system combines a structured architecture candidate generator with an LLM Stack using retrieval-augmented generation and chain-of-thought prompting, plus an iterative feedback loop. Evaluation demonstrates end-to-end synthesis and execution of three accelerator designs on real FPGA hardware, with results showing the approach captures kernel-specific compute/memory trade-offs while reducing manual design effort.