> That already makes strong modeling assumptions (usually including IID, Gaussian noise, etc.,) to get the parameter estimates in the first place

You lose me here - I don't agree with "usually". I guess you're thinking of examples where you are sampling from a population and estimating features of that population. There's nothing wrong with that, but that is a much smaller domain than curve fitting in general.

If you give me a set of x and y, I can fit a parametric curve that tries to minimises the average squared distance between fit and observed values of y without making any assumptions whatsoever. This is a purely mechanical, non-stochastic procedure.

For example, if you give me the points {(0,0), (1,1), (2,4), (3,9)} and the curve y = a x^b, then I'm going to fit a=1, b=2, and I certainly don't need to assume anything about the data generating process to do so. However there is no concept of a confidence interval in this example - the estimates are the estimates, the residual error is 0, and that is pretty much all that can be said.

If you go further and tell me that each of these pairs (x,y) is randomly sampled, or maybe the x is fixed and the y is sampled, then I can do more. But that is often not the case.

For example, suppose you measure the decay of a radioactive source at fixed times t = 0,1,2,... and fit y = A e^{-kt}. The only randomness is small measurement error with, say, SD = 0.5. The bootstrap sees the huge spread in the y-values that comes from the deterministic decay curve itself, not from noise. It interprets that structural variation as sampling variability and you end up with absurdly wide bootstrap confidence intervals that have nothing to do with the actual uncertainty in the experiment.

However the embedding dimension sets the rank of the token representation space. Each layer can transform or refine those vectors, but it can’t expand their intrinsic capacity. A tall but narrow network is bottlenecked by that width. Width-first scaling tends to outperform pure depth scaling, you want enough representational richness per token before you start stacking more layers of processing.

So yeah, embedding size doesn’t have to scale up in lockstep with model size, but in practice it usually does, because once models grow deeper and more capable, narrow embeddings quickly become the limiting factor.

We used reverse geocoding extensively — but never down to street addresses, always to a higher level. We wanted to split measurements by country, region, city — any geographic unit. When you deal with country borders, you get a lot of weird measurements as phones roam onto foreign networks. We weren’t interested in reporting on the experience of users roaming while abroad, so we needed shapefiles good enough to filter all that out and to partition the rest of the data cleanly.

We built a 30-machine Spark cluster on AWS back when Spark was still super early — around v0.7, definitely before 1.0. At the time, you pretty much had to use Scala with Spark if you cared about performance. Most of the workload was point-in-polygon tests. Before that, we were using a brutally hacky pipeline involving PostGIS, EMR, and Pig, and it was hell.

It was incredibly fun, but looking back now, I can see so clearly all the mistakes I made.

Comments URL: https://news.ycombinator.com/item?id=43184932

Points: 5

# Comments: 2

If I wanted to run a production-grade service using the full Deepseek model, with good tokens/sec and the ability to serve concurrent requests, what sort of hardware are we looking at?

I'm not suggesting this is actually a problem, but that's how an argument could go.

Comments URL: https://news.ycombinator.com/item?id=42451624

Points: 2

# Comments: 0

Job seekers almost never actually know if the job was real or not, so it's hard to see how Glassdoor reviews can ever provide the insight this work is looking for.

I do believe that "ghost" jobs exist, often for H1B purposes, but I don't think this work proves it.

Whatever approach you take, you'll probably be selecting the final set by hand, so it's just about building the candidate set in an efficient manner. Low absolute accuracy isn't really an issue as long as you end up with a managable set to review.

I really don't think that is a controversial position, but clearly you disagree.

I have been to Denmark many times and it strikes me as considerably more homogenous than the UK or the US.