时间： 2023.8.8-2023.8.14
最新技术：
LLMs as Database Administrators
论文：arxiv.org
Political Biases Found in NLP Models
Evaluating LLMs as Agents
论文：arxiv.org
Studying LLM Generalization with Influence Functions
论文：arxiv.org
Seeing Through the Brain
论文：arxiv.org
SynJax
论文：arxiv.org
Synthetic Data Reduces Sycophancy in LLMs
论文：arxiv.org
Photorealistic Unreal Graphics (PUG)
论文：arxiv.org
LLMs for Industrial Control
论文：arxiv.org
Trustworthy LLMs
论文：arxiv.org
Automatically Correcting Large Language Models: Surveying the landscape of diverse self-correction strategies
论文： arxiv.org
UniversalNER：在LLM中蒸馏实现开放NER
TPTU：任务规划和工具使用的LLM Agents
论文：arxiv.org
Training Data Protection with Compositional Diffusion Models
论文：arxiv.org
Generative Agents: Interactive Simulacra of Human Behavior
论文：arxiv.org
代码：github.com
课程：
TRL 正式推出，来训练你的首个 RLHF 模型吧！
Fine-Tuning Llama-2: A Comprehensive Case Study for Tailoring Models to Unique Applications
商业：
GPT-4.5前瞻：代码解释器，编程新纪元
LLM-Native产品的变与不变
OpenAI公布「官方爬虫」：GPT-5靠它训练
案例：
DXL-controlnet: OpenPose (v2)
XLANG Agents
facechain
Llama-2-Onnx
candle