Top all-time · #13223
Bojie Li
Personal blog · 01.meaiscience
Bojie Li's personal blog at 01.me covers ai and science. Since 2026, 6 different Hacker News readers have linked it 6 times, most often “Incompressible Knowledge Probes: Measuring what frontier models know”. It is still writing: the last post was on 2026-06-26, and you can follow it by feed.
Too few sampled comments (2) to say how often HN recommends it.?
Why HN links it
HN readers link Bojie Li's research on estimating the parameter counts of frontier LLMs from the models' knowledge, used to check claimed model sizes such as Grok and Opus.
Summarised by Claude from a sample of 2 HN comments that link here. Each comment is marked recommending, critical or neutral (citing without judging); how this works.
Personal blog of Bojie Li
Most-linked posts?
- Incompressible Knowledge Probes: Measuring what frontier models know5 people · 5 mentions · first linked 2026-04-29
What Hacker News says?
> While this source's reliability is certainly debatable Massive understatement. Nowadays it has become hard to find a single Musk statement that doesn't contain at least one lie. > the size matches the results of this paper, in which researchers estimated the parameter count from model knowledge. https://01.me/research/ikp/ Thanks for the pointer. This estimation has Grok 6 times bigger than Musk claims it is, so m…
Comment linking /research/ikp/
on “Outsourcing plus LocalAI will soon become more economical vs. Frontier labs”Submitted Incompressible Knowledge Probes: Measuring what frontier models know as a story · 3 points
HN discussion> What's your source for Opus being a 5T model? Elon Musk tweeted that Grok is 0.5T or 1/10th the size of Opus. https://xcancel.com/elonmusk/status/2042123561666855235#m While this source's reliability is certainly debatable, the size matches the results of this paper, in which researchers estimated the parameter count from model knowledge. https://01.me/research/ikp/
Comment linking /research/ikp/
on “Outsourcing plus LocalAI will soon become more economical vs. Frontier labs”Submitted NameRank: The Model Knows Your Project, Not You as a story · 1 points
HN discussionSubmitted Measuring what frontier models know – and how big they are as a story · 1 points
HN discussionSubmitted Incompressible Knowledge Probes: Measuring Frontier LLM Sizes as a story · 1 points
HN discussion