Pinned Curious about how Replit Agent works? At @Replit we hosted tech talks discussing agent's internals. Let's dive into what goes into an LLM that powers the agent - slides from my talk. 🧵
Anyone remember Bard? Is it another plugin for ChatGPT? Here is the launch week search interest with banana for scale.
How does OpenAI plugins/Browser work? A thread of detailed analysis on the server interaction. S/o to @CrisGiardina for helping with access. TLDR: Browsing (and possibly plugin) is a different model with 8k seq length support and a toolformer like operation structure.
Recently got my O1 visa stamped 🇺🇸🚀 Excited to continue building @Replit Agent - the real extraordinary alien
It looks like @OpenAI has already shipped the required front-end code (including APIs) for loading and executing plugins for everyone.
Randomly bumping into @karpathy at a dinner spot is peak SF.
. @langchain conversation agents can easily consume @OpenAI plugins when given the URL. Here is one example for @Klarna using their hosted API information: klarna.com/.well-known/ai…It should be easy to onboard others plugins even when they require authentication.
Microsoft and Google will kill so many GPT wrapper startups in the coming days.
The best @OpenAI plugins will not be programmatic automation tools but freelance service providers like @Taskrabbit or @fiverr . Long context length and retrieval support would allow LLMs to orchestrate workers towards real-world tasks - using humans as an API to the real world.
Bing Jailbreak: The new Bing search is susceptible to token-smuggling attack. We can get it to generate output for a prompt of adversaries choice! Here is my first attempt at tricking the system to generate malicious output (discretion is advised). #Microsoft#Bing#jailbreak
Replying to @CrisGiardina Wonder what would happen if you play this scene against it.
This flew under the radar for some reason. The key insight is that simply training on synthetic data with positive pairs (correct problem-solution pairs) can lead to the model learning spurious correlations; creatively introducing negative samples makes the process efficient.
Was curious about the basic arithmetic abilities of Claude 3 Opus against GPTs. Designed a tiny experiment over the weekend and the results are surprising. Opus is much better than GPTs with numbers! Let's dive into the results 🧵
