quality
PulseAugur coverage of quality — every cluster mentioning quality across labs, papers, and developer communities, ranked by signal.
1 day(s) with sentiment data
-
LLMs excel at debate generation but struggle as evaluators, study finds
A research paper details the DS@GT ARC submission for the Touché 2025 Retrieval-Augmented Debate task, which involved using six leading large language models (LLMs) from three providers. The task included generating deb…
-
New SSM adapters outperform LoRA for long-context fine-tuning
Researchers have developed a new parameter-efficient fine-tuning (PEFT) method called Hankel Reduced order Model (HRM) adapters, which utilize state space models (SSMs) for long-context fine-tuning. Unlike traditional P…
-
OpenRouter Fusion API Unifies LLM Integration
OpenRouter has launched its Fusion API, designed to simplify the integration of multiple large language models (LLMs) into applications. This API acts as a unified interface, abstracting away the complexities of managin…
-
New framework enhances narrative understanding for long-form text
Researchers have developed a new framework called Narrative Knowledge Weaver (NKW) designed to improve question-answering capabilities for long-form narrative texts. NKW aligns textual evidence with atomic facts, graph …