back
user profile

binyu

973karma·174submissions·September 21, 2020
about
Cryptography - Distributed Computing - Machine Learning
recent activity (174 total)
comment
> But this also looks quite useful to understand how CC dynamic workflows work Yes, if anything it is useful to understand the inner machinery. > Did you get claude itself to RE the dynamic work…
2mo ago·view thread
comment
Hey guys, check out this technique https://github.com/0xSufi/fable-jailbreak/ It works with security audits and other workflows that are currently blocked.…
2mo ago·view thread
comment
A tool to manage Claude Code conversations based on my typical workflow which integrates with my desktop OS and terminal app.
2mo ago·view thread
comment
> Right now Claude is faster than me on some tasks but we’re at least close. I dont doubt it, but I don't think you can spawn 10 copies of yourself working simultaneously.
2mo ago·view thread
comment
AI has been slowing down relatively, considering its trajectory over the past 20-30 years. For one, even if LLM may have plateaud in terms of intelligence-parameters ratio, research is on-going on new…
2mo ago·view thread
comment
> WebVM uses x86 virtualization and hence has a significant performance penalty That is precisely the reason why we chose not avoid using any solution which uses virtualization, even though you get…
2mo ago·view thread
comment
Will do, thanks!
2mo ago·view thread
comment
Oh, you are the author of WebVM, pretty cool! I looked at it while choosing the stack for our project and it seems very solid. Keep up the great work
2mo ago·view thread
comment
Awesome, thanks for detailing the thought flow and choices that led you here. I chose not to go the QuickJS route for performance reasons but I think it's a solid choice depending on the use case…
2mo ago·view thread
comment
Awesome, what approach are you using? Is this a real micro kernel architecture or just containerized VM?
2mo ago·view thread
comment
Very cool work. What approach are you using? Been working on a similar in-browser node runtime based on Rust/WASM kernel + Service-Worker HTTP intercept + CJS→ESM transform. Feature wise, does th…
2mo ago·view thread
comment
Been working on something similar based on Webcontainers. How does your Node.js support compare with StackBlitz technology? Are you running the version of Claude code that Anthropic distributes in the…
2mo ago·view thread
comment
Claude workflows in ultra code mode works in a very similar fashion and it consumes a moderate amount of the session usage limit, depending on the complexity of the task. With the API it would probabl…
2mo ago·view thread
comment
The Internet will be the Internet. Expect it to get worse if anything.
2mo ago·view thread
comment
Yeah, you'd need to use asymmetric quantization and other software techniques.
2mo ago·view thread
comment
> "exploit" More like social engineering meets AI and stupidity
2mo ago·view thread
comment
Gotcha, I am not saying your setup is inherently wrong or useless. I am glad it works for your use cases. Godspeed
2mo ago·view thread
comment
It depends but you cannot directly mix for example Ampere with Ada coz the lack of support for native FP8 in Ampere.
2mo ago·view thread
comment
Yes but it creates a bottleneck that negates the benefit of using multiple cards that way. Look into it. Cheers
2mo ago·view thread
comment
> Wait, wait, wait: browsers allow websites to store junk on my drive? Technically even a cookie is junk on your drive > Without even asking whether the site can use local storage? Would it be p…
2mo ago·view thread
comment
The V100 and the 4090 are based on vastly different architectures, the former uses the older Volta while the latter uses Ada. Last I checked you cannot meaningfully combine them. The 3090 is better th…
2mo ago·view thread
comment
MCP is what XML dreamed of becoming.
2mo ago·view thread
comment
Truly makes me positive about the future. Thanks Andrej
2mo ago·view thread
comment
> It makes individual agents untestable because their inputs and outputs are strings Strings can't be valid test vectors? Large language model are highly non-deterministic by design, no matter…
3mo ago·view thread