Sabitlenmiş Tweet
Adriano
212 posts
Adriano
@jld_adriano
finding beauty in the latent space code @exaailabs (hiring, don't be a stranger)
sf Katılım Haziran 2023
1.8K Takip Edilen295 Takipçiler
@fine_whines so you are saying the delta is from base model performance and not from better search tool use
English
@jld_adriano Better search can naturally lead to better learning in RLVR setups because fewer rollouts totally fail and have zero advantages.
But in particular, my hunch is that Exa search degrades more gracefully and has less variance for slightly different query formulations
English
Fun nugget from our research here -- using a better search backend during training leads to better performance at inference time not just on the one it was trained on, but also when you swap the backend for each agent.
Exa@ExaAILabs
How does Exa compare to Google for training LLMs to search? In this blog post, we find that LLMs using Exa during reinforcement learning reach higher performance with 70% less training compute. exa.ai/blog/rl-search…
English
so i am…. a b2b saas??
Marc Andreessen 🇺🇸@pmarca
“A man is what he thinks about all day.”—Emerson
English
Adriano retweetledi
superpowers
Exa@ExaAILabs
Introducing Exa for Claude: Give Claude access to billions of websites, docs, papers, people, companies, and more. Exa turns Claude into a recruiter, a salesperson, a researcher, a reliable engineer, all with one plugin. Run in terminal: claude plugin marketplace update claude-plugins-official && claude plugin i exa@claude-plugins-official && claude "call mcp__plugin_exa_exa__authenticate (a deferred tool available to you, guaranteed)"
English
what it takes to scale a product’s internal and external complexity
Exa@ExaAILabs
A search engine has many moving parts, and they constantly change with scale of index and traffic. In this blog post, we dive into the code-level abstractions we've built to develop search infrastructure that is both performant and a joy to work with. exa.ai/blog/composing…
English
Adriano retweetledi
We're excited to partner with @coinbase to enable agents to natively pay for web search, via x402!
x402 is an open protocol that enables agents to pay via HTTP, governed by the Linux Foundation. When an Exa API request is made without an API key, Exa now returns a 402 status code with payment information that an agent can act on.
English
Adriano retweetledi
What does it take to store the web as a database?
exa-d is our internal data framework that orchestrates declarative typed dependencies, sparse updates with precise granularity, efficient and parallel execution across scaling compute, and more.
exa.ai/blog/exa-d
English
Adriano retweetledi
Introducing state-of-the-art People Search:
You can now semantically search over 1 billion people using a hybrid retrieval system backed by finetuned Exa embeddings.
Try it: exa.ai
We also created an eval: exa.ai/blog/people-se…
English
@jld_adriano Exa wouldn’t be the same without you - so glad you’re here!
English
@ChangJonathanC @jeffzwang yeah frontend is a bit buggy, we will be streamlining soon
English
@jld_adriano @jeffzwang oh i just tried it, i think there's a bug that made me believe i was using fast mode but it wasn't
English
One of the most insane engineering efforts I've ever seen
Exa@ExaAILabs
Today we're releasing the fastest search API in the world. Exa Fast is a streamlined version of our neural/keyword search with avg latency below 425ms. Link to try it below 👇
English
here is a secret:
most web search apis are a wrapper on top of google.
they have server farms that take user queries, process them in google, and serve the results:
this takes over 700ms
but because we built our tech from scratch (crawl the web, train models, own vector database) we can do that in just
423ms
Exa@ExaAILabs
Today we're releasing the fastest search API in the world. Exa Fast is a streamlined version of our neural/keyword search with avg latency below 425ms. Link to try it below 👇
English