5 ms·

I might be mistaken, but my understanding from having played around with LangChain for a couple months is that because you’ve got to keep all your state in the

I might be mistaken, but my understanding from having played around with LangChain for a couple months is that because you’ve got to keep all your state in the context window, giving the model access to a vectorstore containing the entire chat history allows it to retrieve relevant messages against your query that can then be stuffed or mapreduced into the context window for the next response. The alternative - and I believe the way the ChatGPT web app currently works - is just to stuff/mapreduce user input and machine response into the context window on each step, which quickly gets quite lossy.

kordlessagain 3y ago
You aren't mistaken. Keeping state, or storing memories, is where it's at with prompts. The trick is knowing what to remember and what to forget. I consider vector engines to be "hot" models, given they are storing the vector representations of text already run through the "frozen" model. Having written something a while back that indexes documents and enters into discussion with them, I'm pretty sure ChatGPT is using some type of embedding lookup/match/distance on the history in the window. That means not all text is submitted at the next entry, but whatever mostly matches what is entered by the user (in vector space) is likely pulled in and sent over in the final prompt.
- danielmarkbruce 3y ago
  Sure - but a vector db is helping you keep your prompts to under size X. It isn't adding state and there are various mechanisms to keep your prompt to under size X - like summarization, providing a table of contents etc. It seems to me that vector db and semantic search are one trick in a pile of tricks to keep prompt sizes down until we can get the input sizes up (although gpt4 already takes 32,000 tokens). Using semantic search to find relevant chunks seems misguided but practical in the short term. One of the key benefits of LLMs is they can take into account a lot of context.
  - darkteflon 3y ago
    Context constraint is a cheap way to keep the model on-topic. So rather than relying on an ever-growing context window to stuff/mapreduce more undifferentiated “context” (the entire chat history), interposing a vector search engine that only returns relevant context tends to get you better overall model performance, in addition to being scalable in a way that increasing context window size is not.
    - danielmarkbruce 3y ago
      Agreed. But summarization is better to keep the model on topic for most cases. And there are other tricks. Vectors and semantic search are one (likely questionable way given LLMs can likely reason over a table of contents or similar better) to search a large corpus or very large document. It's really only appropriate for a specific set of use cases. It's not some "general memory layer" for AI.
      - appenz 3y ago
        Summarization is much more expensive than vector db's. Assume you have 1m tokens of context. You could run all through GPT-4 and summarize the information, but it would cost $60 (based on current prices) and take 10's of minutes of GPU time to do the inference. Disclaimer: I work for a16z and on the infra team, so consider me biassed.
        danielmarkbruce 3y ago
        If you look through the comments here, folks are mostly referring to keeping for example a chat history. No one is doing 1m words of chat. A common pattern is to summarize a chat history and pass that in the prompt. As for a corpus of documents (which is what you are presumably talking about), there are a couple problems with what you are saying: First, you are implying that the content is always new - that's not true for many cases folks are talking about solving (like technical support or customer support), so it's a one time fee to summarize the corpus. You might run it periodically for updates. Second, there is an assumption that a basic semantic search is the best way to search documents to find the most relevant content. That's questionable before the existence of LLMs, but with LLMs you are basically assuming your cosine similarity search on your vectors is better than an LLM can do with a simple table of contents and question "where should I search?" I haven't seen someone do a detailed study, but the implicit assumption that semantic search is the best idea for text could easily be a bad one. Third, it assumes the quantum of data to search through is astronomically large and/or getting bigger compared to almost certain decreases in inference cost and increases in input tokens. This will be true for some subset of things, but unlikely to be many and in the cases it is true they'll do something more sophisticated than embeddings and embedding search. They'll probably fine tune the underlying model on an ongoing basis. Regardless - the post you guys wrote seems... like a stretch for a definition of what this really is And, at least on the surface vector databases appear to be commodity infra. Pinecone might be growing fast now, but how do they ever make much money above their costs? But, you guys seem smart, so maybe there is something there?
rahimnathwani 3y ago
Yes, one of the memory options in Langchain is to use a vector store: https://python.langchain.com/en/latest/modules/memory/types/vectorstore_retriever_memory.html https://python.langchain.com/en/latest/modules/memory/types/... It also has a more basic version that just keeps a log of past messages. I don't know whether there's a way (or even a need) to combine these approaches. In a long conversation, it might be useful to trust more recent information more than earlier messages, but Langchain's vector memory doesn't care about sequence.

kordlessagain 3y ago

You aren't mistaken. Keeping state, or storing memories, is where it's at with prompts. The trick is knowing what to remember and what to forget. I consider vector engines to be "hot" models, given they are storing the vector representations of text already run through the "frozen" model. Having written something a while back that indexes documents and enters into discussion with them, I'm pretty sure ChatGPT is using some type of embedding lookup/match/distance on the history in the window. That means not all text is submitted at the next entry, but whatever mostly matches what is entered by the user (in vector space) is likely pulled in and sent over in the final prompt.

danielmarkbruce 3y ago
Sure - but a vector db is helping you keep your prompts to under size X. It isn't adding state and there are various mechanisms to keep your prompt to under size X - like summarization, providing a table of contents etc. It seems to me that vector db and semantic search are one trick in a pile of tricks to keep prompt sizes down until we can get the input sizes up (although gpt4 already takes 32,000 tokens). Using semantic search to find relevant chunks seems misguided but practical in the short term. One of the key benefits of LLMs is they can take into account a lot of context.
- darkteflon 3y ago
  Context constraint is a cheap way to keep the model on-topic. So rather than relying on an ever-growing context window to stuff/mapreduce more undifferentiated “context” (the entire chat history), interposing a vector search engine that only returns relevant context tends to get you better overall model performance, in addition to being scalable in a way that increasing context window size is not.
  - danielmarkbruce 3y ago
    Agreed. But summarization is better to keep the model on topic for most cases. And there are other tricks. Vectors and semantic search are one (likely questionable way given LLMs can likely reason over a table of contents or similar better) to search a large corpus or very large document. It's really only appropriate for a specific set of use cases. It's not some "general memory layer" for AI.
    - appenz 3y ago
      Summarization is much more expensive than vector db's. Assume you have 1m tokens of context. You could run all through GPT-4 and summarize the information, but it would cost $60 (based on current prices) and take 10's of minutes of GPU time to do the inference. Disclaimer: I work for a16z and on the infra team, so consider me biassed.
      - danielmarkbruce 3y ago
        If you look through the comments here, folks are mostly referring to keeping for example a chat history. No one is doing 1m words of chat. A common pattern is to summarize a chat history and pass that in the prompt. As for a corpus of documents (which is what you are presumably talking about), there are a couple problems with what you are saying: First, you are implying that the content is always new - that's not true for many cases folks are talking about solving (like technical support or customer support), so it's a one time fee to summarize the corpus. You might run it periodically for updates. Second, there is an assumption that a basic semantic search is the best way to search documents to find the most relevant content. That's questionable before the existence of LLMs, but with LLMs you are basically assuming your cosine similarity search on your vectors is better than an LLM can do with a simple table of contents and question "where should I search?" I haven't seen someone do a detailed study, but the implicit assumption that semantic search is the best idea for text could easily be a bad one. Third, it assumes the quantum of data to search through is astronomically large and/or getting bigger compared to almost certain decreases in inference cost and increases in input tokens. This will be true for some subset of things, but unlikely to be many and in the cases it is true they'll do something more sophisticated than embeddings and embedding search. They'll probably fine tune the underlying model on an ongoing basis. Regardless - the post you guys wrote seems... like a stretch for a definition of what this really is And, at least on the surface vector databases appear to be commodity infra. Pinecone might be growing fast now, but how do they ever make much money above their costs? But, you guys seem smart, so maybe there is something there?

rahimnathwani 3y ago

Yes, one of the memory options in Langchain is to use a vector store: https://python.langchain.com/en/latest/modules/memory/types/vectorstore_retriever_memory.html https://python.langchain.com/en/latest/modules/memory/types/... It also has a more basic version that just keeps a log of past messages. I don't know whether there's a way (or even a need) to combine these approaches. In a long conversation, it might be useful to trust more recent information more than earlier messages, but Langchain's vector memory doesn't care about sequence.