scribu

8 bulan yang laludiscuss

> But isn't it true for JavaScript too?

You're right, the equivalent JS script produces the same sequence of outputs.

It turns out there is a way to emulate Python's asyncio.create_task().

Python:

  await asyncio.create_task(child())

JavaScript:

  const childTask = new Promise((resolve) => {
    setTimeout(() => child().then(resolve), 0)
  })
  await childTask

scribu

8 bulan yang laludiscuss

> Awaiting a coroutine does not give control back to the event loop.

I think this is a subtler point than one might think on first read, which is muddled due to the poorly chosen examples.

Here's a better illustration:

  import asyncio
  
  async def child():
      print("child start")
      await asyncio.sleep(0)
      print("child end")
  
  async def parent():
      print("parent before")
      await child()        # <-- awaiting a coroutine (not a task)
      print("parent after")
  
  async def other():
      for _ in range(5):
          print("other")
          await asyncio.sleep(0)
  
  async def main():
      other_task = asyncio.create_task(other())
      parent_task = asyncio.create_task(parent())
      await asyncio.gather(other_task, parent_task)
      
  asyncio.run(main())

It prints:

  other
  parent before
  child start
  other
  child end
  parent after
  other
  other
  other

So the author's point is that "other" can never appear in-between "parent before" and "child start".

Edit: clarification

scribu

11 bulan yang laludiscuss

ChatGPT Plus has that (used to be in the free tier too). You can toggle between versions for each of your messages with little left-right arrows.

scribu

12 bulan yang laludiscuss

I’m confused by the clustering step:

> To find the most informative examples, we separately cluster examples labeled clickbait and examples labeled benign, which yields some overlapping clusters

How can you get overlapping clusters if the two sets of labelled examples are disjoint?

scribu

tahun laludiscuss

It’s interesting that they decided to move all of the architecture-specific image-to-embedding preprocessing into a separate library.

Similar to how we ended up with the huggingface/tokenizers library for text-only Tranformers.

scribu

tahun laludiscuss

This seems to be a system to generate better prompts to be fed into a base multimodal model.

Interesting, but title is definitely clickbait.

scribu

tahun laludiscuss

You could run the build process with chroot or inside Docker, so that the hardcoded paths actually resolve to a designated subdirectory.

scribu

tahun laludiscuss

From their Notion page:

> Skywork-OR1-32B-Preview delivers the 671B-parameter Deepseek-R1 performance on math tasks (AIME24 and AIME25) and coding tasks (LiveCodeBench).

Impressive, if true: much better performance than the vanilla distills of R1.

Plus it’s a fully open-source release (including data selection and training code).

scribu

tahun laludiscuss

“Fill in the gaps by using context” is the hard part.

You can’t pre-bake the context into an LLM because it doesn’t exist yet. It gets created through the endless back-and-forth between programmers, designers, users etc.

scribu

tahun laludiscuss

The point is that you’d expect a roughly even distribution of clockwise and counterclockwise spins, not all of them to rotate in the same direction.

scribu

tahun laludiscuss

If the base models already have the “reasoning” capability, as they claim, then it’s not surprising that they were able to get to SOTA using a relatively negligible amount of compute for RL fine-tuning.

I love this sort of “anti-hype” research. We need more of it.

scribu

tahun laludiscuss

The install speed alone makes it worthwhile for me. It went from minutes to seconds.

scribu

tahun laludiscuss

The HN submission title is editorialized in a non-helpful way. Why beat a dead horse instead of focusing on what’s actually new in TFA?

The linked paper proposes an obvious-in-retrospect form of data augmentation: shuffle the order of the premises, so that the model can’t rely on spurious patterns. That’s kinda neat.

scribu

tahun laludiscuss

Would be curious to know how this stacks up against Coconut [1] which also uses latent space for reasoning.

[1] https://arxiv.org/abs/2412.06769

scribu

2 tahun yang laludiscuss

The repo contains only the PDF, not actual runnable code for the RL training pipeline.

Publishing a high-level description of the training algorithm is good, but it doesn't count as "open-sourcing", as commonly understood.

scribu

2 tahun yang laludiscuss

Nice approach!

I'm intrigued by the ability to start execution from a particular task.

One thing I like about LangGraph is the declarative state merging. In the MapReduce example, how do you guarantee that the collector.append() operation is thread-safe?

scribu

2 tahun yang laludiscuss

Actually, there have been attempts to do quantized backprop, but not sure how successfully.

scribu

2 tahun yang laludiscuss

I'm working on something like that. Feel free to email me (address in my bio).

scribu

3 tahun yang laludiscuss

I’m not sure about the Speed chart. I would expect gpt-4-turbo to be faster than plain gpt-4.

scribu

3 tahun yang laludiscuss

I understand that it's a different approach, but I would still have expected this paper to at least mention FlashAttention [1] since they both leverage flash memory.

[1] https://arxiv.org/abs/2205.14135

Submissions

Apple Calls Out EU for Contradictory App Store Rules Under DSA and DMA

FormulaOne – new LLM benchmark using Dynamic Programming problems

Robustly Improving LLM Fairness in Realistic Settings via Interpretability

Streaming Brain-to-Voice Neuroprosthesis

Manus probably isn't China's second 'DeepSeek moment'

Ambiguous evidence predictably leads to polarisation

[untitled]

Bolt gives all employees early exercising and loans for stock options

Extracting sensitive information out of GPT2

The Chinese Surveillance State Isn't Effective

On Facebook, Misinformation Is More Popular Now Than in 2016

Foldable Standing Bicycle [video]

The God of Silence Speaks Up

Facebook funnelling readers towards Covid misinformation – study

Facebook removes Trump post over false Covid-19 claim for first time

Functional Institutions Are the Exception

A Tale of Tough Cities

Strategies of a Survivor reality show participant

Marketplaces: Underutilized Fixed Assets

The “zero marginal cost” false narrative

comments