论文标题：Direct Preference Optimization: Your Language Model is Secretly a Reward Model论文链接...
发简信
IP属地：浙江
-
直接偏好优化技术DPO基础理论及推导
-
自适应视图增强的谣言检测图对比学习方法论文标题：Propagation Tree Is Not Deep: Adaptive Graph Contrastive Learning Approach for Rum...
-
生成式大模型的RLHF技术（一）：基础一、概述 大语言模型（LLMs）在预训练的过程中通常会捕捉数据的特征，而这些训练数据通常既包含高质量的也包含低质量的，因此模型有时会产生不被期望的行为，如编造事实，生成有偏见...
-
LoRA：大模型下游任务的低秩适应论文标题：LoRA: Low-Rank Adaptation of Large Language Models论文链接：https://arxiv.org/abs/2106....
-
Megatron-LM：Transformer模型专用分布式张量模型并行方法论文标题：Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallel...
-
思维树：大模型的复杂推理技术论文标题：Tree of Thoughts: Deliberate Problem Solving with Large Language Models论文链接：https:...
-
LIMA：小规模监督数据指令微调论文标题：LIMA: Less Is More for Alignment论文链接：https://arxiv.org/abs/2305.11206[https://arxi...
-
语言模型的自洽性思维链推理技术论文标题：Self-Consistency Improves Chain of Thought Reasoning in Language Models论文链接：https:...
-
GPipe：微批量流水线并行论文标题：GPipe: Easy Scaling with Micro-Batch Pipeline Parallelism论文链接：https://arxiv.org/ab...
-
InstructGPT：语言模型的人类反馈指令对齐
论文标题：Training language models to follow instructions with human feedback论文链接：https://ar...
-
高效底座模型LLaMA论文标题：LLaMA: Open and Efficient Foundation Language Models论文链接：https://arxiv.org/abs/230...
-
TokenGT：Transformer是强大的图学习器论文标题：Pure Transformers are Powerful Graph Learners论文链接：https://arxiv.org/abs/2207.02505...
-
思维链Prompting促进大型语言模型的推理能力论文标题：Chain-of-Thought Prompting Elicits Reasoning in Large Language Models论文链接：https://...
-
GPT-3：大语言模型小样本学习论文标题：Language Models are Few-Shot Learners论文链接：https://arxiv.org/abs/2005.14165[https:/...
-
GPT-2：无监督多任务学习语言模型论文标题：Language Models are Unsupervised Multitask Learners论文链接：https://d4mucfpksywv.cloud...
-
GPT：通用预训练语言模型论文标题：Improving Language Understanding by Generative Pre-Training论文链接：https://www.cs.ubc...