3 ms·

Can people ELI5 what an LLM or AI "Agent" is? Sometimes when I see the term used, it makes sense, and other times there seems to be nothing concrete behind it.

Can people ELI5 what an LLM or AI "Agent" is? Sometimes when I see the term used, it makes sense, and other times there seems to be nothing concrete behind it. [not a comment on OP post]

hall0ween 3y ago
From my understanding, an agent is a collection of tools that can be navigated by an LLM. It follows an approach called reason-action (react) or thought-action-observation. Spit balling here, thought means getting a prompt and having a response to it (i.e., run particular action). An action executes (e.g., executing code in python repl, using a search a api, pulling info from wikipedia). The LLM receives some feedback (observation) and the process starts over.
ZunarJ5 3y ago
Crash course: LLM: https://writings.stephenwolfram.com/2023/02/what-is-chatgpt-doing-and-why-does-it-work/ https://writings.stephenwolfram.com/2023/02/what-is-chatgpt-... Agent: https://lilianweng.github.io/posts/2023-06-23-agent/ https://lilianweng.github.io/posts/2023-06-23-agent/
aldarisbm 3y ago
there’s plenty online, but basically an LLM is a model trained on an expansive corpus of data (unstructured, text) and basically trained to mimic human language. It can answers questions using probabilistic logic. People have been thinking of a way to make LLMs perform actions in the real world, by giving these LLMs “tools”. An LLM with tools that can perform tasks would be an AI Agent
itissid 3y ago
Its a fancy term for s piece of code that has some context(memory) and can communicate with other AI APIs(aka lang chain) including itself. It can retrieve certain information from other sources(called vector stores). What makes it different from other programs is that Agents can do probabilistic inference on the closure natural language, i.e. they can predict(P(Y|X) as well as generate(P(X, y)). Both of these are very hard to do with any one previous model and hence with any service that could combine software API + Classic ML Model(pre Generative+Discriminative models like llm).
localhost 3y ago
I think about agents as being an entity that you can collaborate with. Among other things, it needs 1/ memory, 2/ ability to create and execute long-term plans. One of the best papers that I have read on this topic is the Generative Agents paper by Park [1] Today's LLMs are kind of like Dory from Finding Nemo - you have to recreate the context every time you do a slightly different task, or when the context window for the LLM is no longer sufficient to remember previous turns in a conversation. An agent can sit on the other side of a piece of collaborative software that was designed for human collaboration. This is why chatbots are the current "killer app" for AI, we already understand how to collaborate with other people via chat. Now imagine what we can do with more sophisticated pieces of software like Figma or Excel. Disclosure: I work on the Python in Excel feature and we just announced Copilot for Excel a couple of days ago. [2] Other modalities that excite me are tools like Vision Pro which will be an interesting test bed for multi-modal generative agents. [1] https://arxiv.org/abs/2304.03442 https://arxiv.org/abs/2304.03442 [2] https://twitter.com/john_lam/status/1704886164753363261 https://twitter.com/john_lam/status/1704886164753363261
DougBTX 3y ago
It is frameworks vs libraries. An agent decides when to call your code, or possibly someone else’s code.
jondwillis 3y ago
here- try to my buggy, unfinished demo, you’ll probably understand it: https://waggledance.ai https://waggledance.ai source is available if you tap the GH icon in the footer.
thoughtlede 3y ago
I would characterize an LLM Agent to be "a software application of LLM" that is typically used to solve a problem that requires multiple steps. The application uses an LLM to breakdown the problem into pieces and asks the LLM to choose a tool that is suited to handle each piece. The application then invokes the tool to solve that piece of the problem. This is iteratively done until the original problem is solved or determined to be unsolvable. There are many agent variations, depending on whether you breakdown the problem once at the beginning, or understand what to do next based on the outcome of the previous step, or constantly evaluate and reprioritize which piece of the problem to solve next. Because you rely on an LLM to understand the problem, if you chose an LLM that exhibits good world understanding, and designed the LLM Agent well enough, then you can use it to solve problems that traditionally required deterministic code.
isaacfung 3y ago
LLM has no long term memory(limited by context size or you have to do fine tuning which isnt efficient) and can't interact with the outside world. You can augment it with retrieval systems(vector store, graph db, sql db + lexical/semantic/graph search) and give it access to tools like api, search engine and code interpreter. Then it becomes an agent.
- isaacfung 3y ago
  https://github.com/WooooDyy/LLM-Agent-Paper-List https://github.com/WooooDyy/LLM-Agent-Paper-List
  - zvr 3y ago
    This is literally the HN submission you're commenting on.
    - isaacfung 3y ago
      I am an idiot. I just read the comments without looking at the linked page.
jsemrau 3y ago
The Encyclopedia Autonomica defines Agent as a software that is capable of reasoning, using memory, plan its actions, and using tools. Or in short: Agent= Reasoning + Memory + Planning + Tools
lukebuehler 3y ago
Agents have agency: they do stuff. The way the do it is by calling code or writing code and then executing it. They need some retrieval system too (often called memory) that selects the relevant possible actions for a current context. I wrote this article recently that outlines what a system needs to run agents: https://github.com/lukebuehler/agent-os/blob/main/docs/articles/manifesto.md https://github.com/lukebuehler/agent-os/blob/main/docs/artic...
madrox 3y ago
An agent is a style of prompt that lets LLMs act as reasoning engines. It's also known as the ReAct pattern (which engineers are avoiding using for namespace collision reasions). You can read a good intro example here: https://github.com/brexhq/prompt-engineering#react https://github.com/brexhq/prompt-engineering#react

Can people ELI5 what an LLM or AI "Agent" is? Sometimes when I see the term used, it makes sense, and other times there seems to be nothing concrete behind it. [not a comment on OP post]

hall0ween 3y ago

From my understanding, an agent is a collection of tools that can be navigated by an LLM. It follows an approach called reason-action (react) or thought-action-observation. Spit balling here, thought means getting a prompt and having a response to it (i.e., run particular action). An action executes (e.g., executing code in python repl, using a search a api, pulling info from wikipedia). The LLM receives some feedback (observation) and the process starts over.

ZunarJ5 3y ago

Crash course: LLM: https://writings.stephenwolfram.com/2023/02/what-is-chatgpt-doing-and-why-does-it-work/ https://writings.stephenwolfram.com/2023/02/what-is-chatgpt-... Agent: https://lilianweng.github.io/posts/2023-06-23-agent/ https://lilianweng.github.io/posts/2023-06-23-agent/

aldarisbm 3y ago

there’s plenty online, but basically an LLM is a model trained on an expansive corpus of data (unstructured, text) and basically trained to mimic human language. It can answers questions using probabilistic logic. People have been thinking of a way to make LLMs perform actions in the real world, by giving these LLMs “tools”. An LLM with tools that can perform tasks would be an AI Agent

itissid 3y ago

Its a fancy term for s piece of code that has some context(memory) and can communicate with other AI APIs(aka lang chain) including itself. It can retrieve certain information from other sources(called vector stores). What makes it different from other programs is that Agents can do probabilistic inference on the closure natural language, i.e. they can predict(P(Y|X) as well as generate(P(X, y)). Both of these are very hard to do with any one previous model and hence with any service that could combine software API + Classic ML Model(pre Generative+Discriminative models like llm).

localhost 3y ago

I think about agents as being an entity that you can collaborate with. Among other things, it needs 1/ memory, 2/ ability to create and execute long-term plans. One of the best papers that I have read on this topic is the Generative Agents paper by Park [1] Today's LLMs are kind of like Dory from Finding Nemo - you have to recreate the context every time you do a slightly different task, or when the context window for the LLM is no longer sufficient to remember previous turns in a conversation. An agent can sit on the other side of a piece of collaborative software that was designed for human collaboration. This is why chatbots are the current "killer app" for AI, we already understand how to collaborate with other people via chat. Now imagine what we can do with more sophisticated pieces of software like Figma or Excel. Disclosure: I work on the Python in Excel feature and we just announced Copilot for Excel a couple of days ago. [2] Other modalities that excite me are tools like Vision Pro which will be an interesting test bed for multi-modal generative agents. [1] https://arxiv.org/abs/2304.03442 https://arxiv.org/abs/2304.03442 [2] https://twitter.com/john_lam/status/1704886164753363261 https://twitter.com/john_lam/status/1704886164753363261

DougBTX 3y ago

It is frameworks vs libraries. An agent decides when to call your code, or possibly someone else’s code.

jondwillis 3y ago

here- try to my buggy, unfinished demo, you’ll probably understand it: https://waggledance.ai https://waggledance.ai source is available if you tap the GH icon in the footer.

thoughtlede 3y ago

I would characterize an LLM Agent to be "a software application of LLM" that is typically used to solve a problem that requires multiple steps. The application uses an LLM to breakdown the problem into pieces and asks the LLM to choose a tool that is suited to handle each piece. The application then invokes the tool to solve that piece of the problem. This is iteratively done until the original problem is solved or determined to be unsolvable. There are many agent variations, depending on whether you breakdown the problem once at the beginning, or understand what to do next based on the outcome of the previous step, or constantly evaluate and reprioritize which piece of the problem to solve next. Because you rely on an LLM to understand the problem, if you chose an LLM that exhibits good world understanding, and designed the LLM Agent well enough, then you can use it to solve problems that traditionally required deterministic code.

isaacfung 3y ago

LLM has no long term memory(limited by context size or you have to do fine tuning which isnt efficient) and can't interact with the outside world. You can augment it with retrieval systems(vector store, graph db, sql db + lexical/semantic/graph search) and give it access to tools like api, search engine and code interpreter. Then it becomes an agent.

isaacfung 3y ago
https://github.com/WooooDyy/LLM-Agent-Paper-List https://github.com/WooooDyy/LLM-Agent-Paper-List
- zvr 3y ago
  This is literally the HN submission you're commenting on.
  - isaacfung 3y ago
    I am an idiot. I just read the comments without looking at the linked page.

jsemrau 3y ago

The Encyclopedia Autonomica defines Agent as a software that is capable of reasoning, using memory, plan its actions, and using tools. Or in short: Agent= Reasoning + Memory + Planning + Tools

lukebuehler 3y ago

Agents have agency: they do stuff. The way the do it is by calling code or writing code and then executing it. They need some retrieval system too (often called memory) that selects the relevant possible actions for a current context. I wrote this article recently that outlines what a system needs to run agents: https://github.com/lukebuehler/agent-os/blob/main/docs/articles/manifesto.md https://github.com/lukebuehler/agent-os/blob/main/docs/artic...

madrox 3y ago

An agent is a style of prompt that lets LLMs act as reasoning engines. It's also known as the ReAct pattern (which engineers are avoiding using for namespace collision reasions). You can read a good intro example here: https://github.com/brexhq/prompt-engineering#react https://github.com/brexhq/prompt-engineering#react