Llama2Vec: Unsupervised adaptation of large language models for dense retrieval
PulseAugur coverage of Llama2Vec: Unsupervised adaptation of large language models for dense retrieval — every cluster mentioning Llama2Vec: Unsupervised adaptation of large language models for dense retrieval across labs, papers, and developer communities, ranked by signal.
3 天有情绪数据
-
Research finds truthfulness is inherited across LLM model families
A new research paper explores the preservation of contextual truthfulness across model lineages, finding that truth scores are strongly maintained from foundational large language models (LLMs) to their downstream varia…
-
SelectiveRM framework trains reward models to ignore noisy preferences
Researchers from Zhejiang University, Xiaohongshu, and Peking University have developed SelectiveRM, a novel framework for training reward models in large language models. This method addresses the issue of noisy prefer…
-
New method combats data laundering in LLM training
一篇新研究论文介绍了一种名为合成数据逆转（SDR）的方法，旨在打击大型语言模型（LLM）训练中的数据洗钱行为。数据洗钱涉及转换专有数据以模糊其来源，使权利所有者难以检测未经授权的使用。SDR通过推断未知的洗钱转换并合成模仿洗钱数据的查询来工作，从而增强检测信号。该方法在MIMIR基准测试中得到验证，在增强各种LLM家族和洗钱实践中的数据滥用检测方面显示出了一致的有效性。