News for hackers. Stuff that matters.

jwesleyharding

Show HN: StructEval - a structured output evaluation and comparison tool

https://github.com/jhiker/structeval

I'm sharing a simple utility I worked on to make evaluation of structured outputs from LLMs a bit simpler.For context:When evaluating structured outputs, you often want to composable comparison logic to allow for meaningful comparison across different types of outputs (free text, enums, ints, and all the other json tyeps). You also want to compare arrays as a multisets -- order-agnostic...

( Read More... | no comments | on HN )
Re:Notes on Virtual Threads and Clojure on Friday May 13, 2022 @01:00PM (#31370037)
to clarity just a tad, the Scala compiler does TCO out of the box and the annotation is added only to check that the method is in fact so optimizable
Re:Java 18 / JDK 18: General Availability on Tuesday March 22, 2022 @11:31AM (#30767297)
1) network effects, around runtime and core library especially. IMO, for enterprise, runtime is especially sticky. 2) of the JVM languages, Java will (always?) be the easiest to hire for
Re:Wherever oil goes, the stock market goes on Monday February 15, 2016 @01:20PM (#11104952)
>I don't think oil is the cause of the market drop: stocks had been overbought for a while (if you look at historic P/E ratios), so it is natural for it to return to normal.

p/e ratios are historically very low relative to interest rates

Re:Economic Inequality on Saturday January 02, 2016 @06:12PM (#10828483)
Did he read Piketty? It doesn't seem like he's even read it somehow: an awkward oversight if you're looking to contribute to high-level discourse about inequality.

He seems to assume that Ycombinator is a microcosm for the American economy, but this is a burdensome assumption for most who have even a casual interest in the subject.

For those who missed it: Capital, the recent intellectual best-seller on this topic, spent about 500 pages attempting to demonstrate that inherited wealth (NOT human capital) is the dominant force in the economy and that this tendency is only worsening, for fundamental structural reasons (r > g). In other words, we are entering a new gilded age.

Think Trump, not Zuckerburg.

So, I can't understand why Graham doesn't address this. If Piketty is correct, it renders meaningless almost every thing he writes in this piece.