A developer at LogicLoop described a production hybrid-search RAG pipeline that runs BM25 keyword retrieval in Elasticsearch alongside dense vector retrieval in pgvector and fuses normalized scores, r…
MERGE, a two-stage framework from researchers behind arXiv paper 2609.37574v1, ensembles three heterogeneous 7-8B open-source LLMs whose candidate query expansions are synthesized by a larger LLM, imp…
Hybrid retrieval that combines full-text search and vector search gives AI agents better recall than embeddings alone, according to a technical guide on the two methods. The guide states that embeddin…
A developer's technical breakdown argues that combining BM25 keyword search with dense vector retrieval via reciprocal rank fusion, then applying a cross-encoder re-ranker over the top candidates, del…
A developer is building Soil Doctor, an agricultural advisory system whose retrieval-augmented generation pipeline combines semantic embeddings, BM25 lexical search, and cross-encoder reranking over a…
A developer's two-stage evaluation of MCP tool selection found that a small model (gpt-5.4-mini) picks the correct tool from a forced five-tool shortlist 90–97% of the time regardless of description v…
A developer improved retrieval overlap on a legal-contract benchmark from 37.3% to 50.4% by retrieving at the sentence level and then returning the full paragraph each hit came from, using stela chunk…
TypeSafe AI released Jev, a "System One" decision model that returns floating point numbers instead of prose and charges only for input tokens at $0.042 per million, undercutting GPT-5 Nano's $0.05 pe…
A resource-efficient Hybrid RAG pipeline using BM25, BGE-M3, and Cross-Encoder reranking scored 0.8095 on the private test set of the UNLP 2026 Shared Task on Multi-Domain Document Understanding, plac…
JEV, a small instruction-following model built for structured decisions, pushed top-1 retrieval accuracy from a 21% BM25 baseline to 54% when added as a reranker in a demonstrated RAG benchmark, accor…
Researchers developed LegMexIA, an AI-enabled research assistant for the Legislación Mexicana corpus, a 42-volume compilation of legal dispositions spanning more than two centuries. LegMexIA combines …
JEV offers new API users $5 in free credit on signup and prices its structured decision-making model below full LLM rerankers, according to hands-on testing shared by an AI engineer. JEV's cost and la…
A training-free routing policy called the Tri-Metric Router achieved 0% out-of-memory failures and 88.5 ± 4.4% oracle alignment on out-of-distribution holdouts for retrieval-augmented generation on co…
A study of on-device tool routing over 600 Korean and English requests and a catalog of 70 local actions found that abstention, not tool selection, is where a neural component is required, according t…
A developer argues that IBM's promoted 'Chunkless RAG' approach, in which an AI agent navigates parsed document structure instead of retrieving fixed-size embedded chunks, is oversold for real-world p…
A developer argues that vector similarity alone is insufficient for coding-agent memory, because a ranked list of semantically close results is not a decision about whether a past conclusion is safe t…
Aito published beta benchmarks claiming its predictive database matches or beats tuned ML baselines on high-cardinality tasks with zero training time, including 68.5% accuracy on invoice routing accep…
A Medium post in the "AI in Plain English" series argues that BM25 keyword ranking remains essential to retrieval-augmented generation (RAG) systems even as embeddings draw widespread attention. The a…
A developer catalogued a 1,245-object database schema with an LLM, generating natural-language descriptions for each table, and found that retrieval quality dropped sharply rather than improving. The …
Pinecone Systems announced that full-text search is now generally available in Pinecone Database, adding BM25 keyword ranking and text-match filters to the same index as dense vectors, enabling exact …