Find a story
Search Spins
Search titles, summaries, and missing voices across published articles — press releases, announcements, and media coverage.
10 results for “retrieval-augmented generation”
TA-RAG: Tone Awareness as a Design Imperative for Retrieval-Augmented Generation
Researchers propose Tone-Aware RAG (TA-RAG), a conceptual framework that prioritizes communicative alignment—such as readability, stigma-free language, and empathetic framing—alongside factual accuracy in retrieval-augmented generation systems, citing persistent misalignments in tone-sensitive domains like public health peer support.
Aug 10, 2026
NeSyFS: A Neuro-symbolic Fast-Slow Thinking Framework for LLM Agent under Partial Observability
A new neuro-symbolic framework called NeSyFS is proposed to improve LLM agent decision-making under partial observability by integrating fast-reactive and slow-reflective reasoning modules with a knowledge graph–based belief state representation.
Aug 3, 2026
SCAIR: Schema-Conditioned Agentic Iterative Reasoning for Enterprise Knowledge Graphs
Researchers introduced SCAIR, a training-free framework for improving natural language querying over enterprise knowledge graphs by embedding schema-aware structural constraints into iterative reasoning — addressing poor generalization of existing agentic methods on real-world, operationally constrained KGs.
Jul 28, 2026
RIMS: Preference Optimization via Smoothed Multi-pair Aggregation for Small-Scale LLM Retrieval-Augmented Generation
A new preference optimization framework called RIMS improves small-scale language model (SLM) performance in retrieval-augmented generation under noisy evidence conditions by replacing hard preference pair selection with a differentiable smooth aggregation mechanism.
Jul 21, 2026
HG-RAG: Hierarchy-Guided Retrieval-Augmented Generation for Structured Knowledge Graphs
A new RAG framework called HG-RAG introduces hierarchical graph traversal over structured knowledge graphs to improve LLM reasoning on hierarchical, relational, and multi-hop queries — addressing a documented limitation of flat-document RAG systems.
Jul 17, 2026
Agentic AI and Retrieval-Augmented Models in Straight-Through Underwriting
A research paper introduces an experimental multi-agent AI system for straight-through underwriting of small commercial insurance policies, claiming superior performance in complex, information-scarce scenarios compared to single-LLM and naive RAG baselines.
Jul 10, 2026
Healthier LLMs: Retrieval-Augmented Generation for Public Health Question Answering
Researchers extended PubHealthBench to evaluate Retrieval-Augmented Generation (RAG) systems for public health question answering, finding hybrid retrieval improves recall and enables smaller LLMs to match larger ones when grounded in official UK guidance.
Jul 10, 2026
PatchOptic for Shared-State LLM Workflows with Projected Views and Verified Structured Updates
PatchOptic is a new interface for LLM agentic workflows that enforces structured, verified updates to shared state using projected views and patch contracts — addressing the gap between local model edits and global state consistency.
Jul 9, 2026
Distill Where the Student Goes: Teacher-Regularized RL for English-Evidence Cross-Lingual RAG
Researchers propose TR-RAG, a teacher-regularized reinforcement learning method to improve cross-lingual RAG performance when users query in non-English languages but retrieved evidence remains English — addressing language drift and unreliable evidence usage.
Jul 8, 2026
RareDxR1: Autonomous Medical Reasoning for Rare Disease Diagnosis Beyond Human Annotation
RareDxR1 is a new end-to-end large language model for rare disease diagnosis that bypasses human-annotated training data and predefined ontologies, claiming state-of-the-art accuracy on open-domain benchmarks.
Published Jul 2, 2026 · Analyzed Jul 5, 2026