dataviz1000

13 ms·

1. 151.▲
   by dataviz1000 6mo ago

   There are some problems reasoning about it. In the thinking output, it isn't really discrete. Therefore I can't apply series-system reliability formulas. But those are a very good metaphor. I can see in the thinking output [1 + 1
2. 152.▲
   Show HN: Reliably Incorrect – explore LLM reliability with data visualizations(adamsohn.com)

   2 points by dataviz1000 6mo ago|3 comments
3. 153.▲
   by dataviz1000 6mo ago

   I'm having an hard time getting my mind to see this. > Users should re-tune their prompts and harnesses accordingly. I read this in the press release and my mind thought it meant test harness. Then there was a blog post about long r
4. 154.▲
   by dataviz1000 6mo ago

   What do you mean by harness here?
5. 155.▲
   by dataviz1000 6mo ago

   Here is a 3blue1brown short about the relationship between words in a 3 dimensional vector space. [0] In order to show this conceptually to a human it requires reducing the dimensions from 10,000 or 20,000 to 3. In order to get the thinking
6. 156.▲
   by dataviz1000 6mo ago

   Thinking helps the models arrive at the correct answer with more consistency. However, they get the reward at the end of a cycle. Turns out, without huge constraints during training thinking, the series of thinking tokens, is gibberish to h
7. 157.▲
   by dataviz1000 6mo ago

   Have a look at https://github.com/adam-s/agent-tuning . Now, I'm working on developing the evaluation, the part that quantifies the performance of the agent. I'm having a hard time explaining it. You should be
8. 158.▲
   by dataviz1000 6mo ago

   I built the same thing as this just for websites. [0] I'm more interested in using Claude recursion to tune itself -- the agent writes itself, the agent -- than hacking websites. It is a good demonstration that 47 iterations of a recur
9. 159.▲
   by dataviz1000 6mo ago

   > to harden a system you need to spend more tokens discovering exploits than attackers will spend exploiting them. I, for the NFL front offices, created a script that exposed an API to fully automate Ticketmaster through the front end so
10. 160.▲
    by dataviz1000 6mo ago

    The question should be did Live Nation knowingly allow scalpers (aka ticket brokers) to corner the market on highest demand events AND create artificial scarcity by only posting a small handful of the tickets they controlled at extreme infl
11. 161.▲
    by dataviz1000 6mo ago

    Because these are probabilistic machines, they solve the same problem at a predictable rate. Even with different variables, the success rate stays consistent. I only noticed the premature quitting issue recently and haven't tested it m
12. 162.▲
    by dataviz1000 6mo ago

    > Anthropic says adaptive thinking decides whether/how much to use extended thinking based on request complexity, with effort as soft guidance and max_tokens as the hard cap Nothing I said contradicts this. Here is the first attemp
13. 163.▲
    by dataviz1000 6mo ago

    That's awesome! You might be interesting in what I've been working on. I've discovered giving an autoresearch approach to letting Claude write Claude, it will find lots of alpha everywhere beating SPY buy and hold. It will ev
14. 164.▲
    by dataviz1000 6mo ago

    I know why. Several people had problems with Sonnet burning through all their credits grinding on a problem it can't solve. Opus fixes this — it has a confidence threshold below which it exits the task instead of grinding. "I spen
15. 165.▲
    by dataviz1000 6mo ago

    I've been working on proving that Claude Opus can be self-reflecting meaning that its attention and context are large enough that it is aware of its own instructions, aware of the task, and capable of writing its own instructions to o
16. 166.▲
    by dataviz1000 6mo ago

    > Claude has gotten noticeably worse for me too. It goes into long exploration loops for 5+ minutes even when I point it to the exact files to inspect. This is what I'm working on proving now. It is more that there is a confidence s
17. 167.▲
    by dataviz1000 6mo ago

    I’ve been traveling around the world. It is 50 / 50 of the socket is properly grounded —-anywhere in the world. I get a tingling zap on the wrist when not properly grounded. The charger also gets hot and sparks.
18. 168.▲
    by dataviz1000 6mo ago

    This is discussed in the Linus vs Linus interview, "Building the PERFECT Linux PC with Linus Torvalds". [0] [0] https://youtu.be/mfv0V1SxbNA?si=CBnnesr4nCJLuB9D&t=2003
19. 169.▲
    Show HN: Agent Tuning, using recursion to achieve predictable agent output(github.com)

    5 points by dataviz1000 6mo ago|0 comments
20. 170.▲
    by dataviz1000 6mo ago

    My next exploration will be "Coding Agents: fail slow, silent, and deceivingly". After one month working on using Claude to create trading strategies, the one thing I learned; if the strategy looks like it can profit, it is a lie.
21. 171.▲
    by dataviz1000 6mo ago

    Sorry to spam, I'm working on this also from a different angle. Hopefully sharing adds to the conversation. First, about the loop, Claude's (coding agent) context and attention is big enough to self-reflect. Agent Tuning shows a t
22. 172.▲
    by dataviz1000 6mo ago

    If you want to benchmark, consider this https://github.com/adam-s/testing-claude-agent
23. 173.▲
    by dataviz1000 6mo ago

    I need to revise what I wrote. The protests and stance have all been against selective service for both men and woman. However, on the flip side, the stance for enlisting and volunteering are opposite. I'll let you Google that one to s
24. 174.▲
    by dataviz1000 6mo ago

    This will end with the only way to authenticate the people we're talking to is meeting them at the coffeeshop in the morning.
25. 175.▲
    by dataviz1000 6mo ago

    Being able to serve is something the feminists have been fighting for the hardest over decades. The people who are trying to make young men only doing the killing the norm are the same people trying to end feminism. Therefore, there is some
26. 176.▲
    by dataviz1000 7mo ago

    Location: Lima, Peru (I'm a US citizen traveling the world for 2 years) Remote: Yes Willing to relocate: Yes Technologies: TypeScript, JavaScript, PHP, Python, FastAPI, Redis, MongoDB, PostgreSQL, TimescaleDB, MySQL, React, React Nativ
27. 177.▲
    by dataviz1000 7mo ago

    I made a test [0] which runs several different configurations against coding tasks from easy to hard. There is a test which it has to pass. Because of temperature, the number of tokens per one shot vary widely with all the different configu
28. 178.▲
    by dataviz1000 7mo ago

    Oh, so the word 'gate' is probably in the documentation also! I see. So this isn't as scary. Claude is helping me understand how to use it properly.
29. 179.▲
    by dataviz1000 7mo ago

    Did you call it '/handoff' or did Claude name it that? The reason I'm asking is because I noticed a pattern with Claude subtly influencing me. For example, the first time I heard the the word 'gate' was from Cl
30. 180.▲
    by dataviz1000 7mo ago

    Here is a description of the iteration loop. [0] I'm working on another draft that will be much more polished and have better explanations of the iteration loop. > There is no proof, just a self-congratulatory word salad with dubiou