Merge branch 'main' into art-zh

pages/techniques/_meta.zh.json

@@ -5,7 +5,7 @@
     "consistency": "自我一致性",
     "knowledge": "生成知识提示",
     "tot": "Tree of Thoughts",
-    "rag": "Retrieval Augmented Generation",
+    "rag": "检索增强生成 (RAG)",
     "art": "自动推理并使用工具（ART）",
     "ape": "自动提示工程师",
     "activeprompt": "Active-Prompt",

pages/techniques/rag.zh.mdx

@@ -1,3 +1,25 @@
-# Retrieval Augmented Generation (RAG)
+# 检索增强生成 (RAG)
 
-This page needs a translation! Feel free to contribute a translation by clicking the `Edit this page` button on the right side.
+import {Screenshot} from 'components/screenshot'
+import RAG from '../../img/rag.png'
+
+通用语言模型通过微调就可以完成几类常见任务，比如分析情绪和识别命名实体。这些任务不需要额外的背景知识就可以完成。
+
+要完成更复杂和知识密集型的任务，可以基于语言模型构建一个系统，访问外部知识源来做到。这样的实现与事实更加一性，生成的答案更可靠，还有助于缓解“幻觉”问题。
+
+Meta AI 的研究人员引入了一种叫做[检索增强生成（Retrieval Augmented Generation，RAG）](https://ai.facebook.com/blog/retrieval-augmented-generation-streamlining-the-creation-of-intelligent-natural-language-processing-models/)的方法来完成这类知识密集型的任务。RAG 把一个信息检索组件和文本生成模型结合在一起。RAG 可以微调，其内部知识的修改方式很高效，不需要对整个模型进行重新训练。
+
+RAG 会接受输入并检索出一组相关/支撑的文档，并给出文档的来源（例如维基百科）。这些文档作为上下文和输入的原始提示词组合，送给文本生成器得到最终的输出。这样 RAG 更加适应事实会随时间变化的情况。这非常有用，因为 LLM 的参数化知识是静态的。RAG 让语言模型不用重新训练就能够获取最新的信息，基于检索生成产生可靠的输出。
+
+Lewis 等人（2021）一个通用的 RAG 微调方法。这种方法使用预训练的 seq2seq 作为参数记忆，用维基百科的密集向量索引作为非参数记忆（使通过神经网络预训练的检索器访问）。这种方法工作原理概况如下：
+
+<Screenshot src={RAG} alt="RAG" />
+图片援引自: [Lewis et el. (2021)](https://arxiv.org/pdf/2005.11401.pdf)
+
+RAG 在 [Natural Questions](https://ai.google.com/research/NaturalQuestions)、[WebQuestions](https://paperswithcode.com/dataset/webquestions) 和 CuratedTrec 等基准测试中表现抢眼。用 MS-MARCO 和 Jeopardy 问题进行测试时，RAG 生成的答案更符合事实、更具体、更多样。FEVER 事实验证使用 RAG 后也得到了更好的结果。
+
+这说明 RAG 是一种可行的方案，能在知识密集型任务中增强语言模型的输出。
+
+最近，基于检索器的方法越来越流行，经常与 ChatGPT 等流行 LLM 结合使用来提高其能力和事实一致性。
+
+LangChain 文档中可以找到[一个使用检索器和 LLM 回答问题并给出知识来源的简单例子](https://python.langchain.com/en/latest/modules/chains/index_examples/vector_db_qa_with_sources.html)。

pages/techniques/tot.en.mdx

@@ -31,3 +31,13 @@ Code available [here](https://github.com/princeton-nlp/tree-of-thought-llm) and
 
 At a high level, the main ideas of [Yao et el. (2023)](https://arxiv.org/abs/2305.10601) and [Long (2023)](https://arxiv.org/abs/2305.08291) are similar. Both enhance LLM's capability for complex problem solving through tree search via a multi-round conversation. One of the main difference is that [Yao et el. (2023)](https://arxiv.org/abs/2305.10601) leverages DFS/BFS/beam search, while the tree search strategy (i.e. when to backtrack and backtracking by how many levels, etc.) proposed in [Long (2023)](https://arxiv.org/abs/2305.08291) is driven by a "ToT Controller" trained through reinforcement learning. DFS/BFS/Beam search are generic solution search strategies with no adaptation to specific problems. In comparison, a ToT Controller trained through RL might be able learn from new data set or through self-play (AlphaGo vs brute force search), and hence the RL-based ToT system can continue to evolve and learn new knowledge even with a fixed LLM.
 
+[Hulbert (2023)](https://github.com/dave1010/tree-of-thought-prompting) has proposed Tree-of-Thought Prompting, which applies the main concept from ToT frameworks as a simple prompting technique, getting the LLM to evaluate intermediate thoughts in a single prompt. A sample ToT prompt is:
+
+```
+Imagine three different experts are answering this question.
+All experts will write down 1 step of their thinking,
+then share it with the group.
+Then all experts will go on to the next step, etc.
+If any expert realises they're wrong at any point then they leave.
+The question is...
+```