Tag
The article introduces RENDER, a benchmark that controls reader-facing evidence in LLM memory evaluations, showing significant accuracy variations based on how memory entries are rendered to the model.
This paper introduces TRIAD, a three-stage automated approach to generate validated datasets for evaluating multi-hop RAG systems in domain-specific settings, assessed against established benchmarks.
LLM-FACETS is an open-source evaluation framework designed to help practitioners assess LLM transparency and accountability with a focus on privacy and data flow transparency. It provides a browser interface, plugin architecture, and supports multiple auditing mechanisms including token-level log-probability visualization and RAG Triad metrics.
This paper presents a comparative evaluation of embedding models and generator backends for Khmer-language retrieval-augmented question answering in the telecom domain, finding that BGE-M3 performs best for retrieval while generator strengths vary across metrics.
RARE introduces a redundancy-aware retrieval evaluation framework that decomposes documents into atomic facts to create realistic benchmarks for high-similarity corpora like finance, legal, and patents, revealing significant performance drops in existing retrievers.