pyentropy

Please. This is IAS -- the members can barely be bothered to commute.

> Hadamard matrix of order 668

Solved? https://epoch.ai/frontiermath/open-problems/hadamard

Not a Lean expert but some of the proof tactics used to prove are probably novel? Or, you could prompt agents later to analyze which lemmas or parts of the proof are surprising or applicable to other problems?

A lot of good repos (CLIs, frameworks) had 'tree' unicode directory structure with like ├──, └──, and │ , as well as emojis for passing/failed tests and README docs, some even had ascii diagrams, maybe a unicode arrow or two, but LLMs absolutely overuse it.

I don't know why chatbots prefer → over -> so much.

This is a HN comment-shaped string, gated with a GitHub scope :-)

If they are targeting arithmetic circuit complexity, it doesn't take a genius to predict what their long term aim is... A permanent is much more than a determinant with plus signs...

Vedran Miletić doesn't sound German to me! :-)

In times when software eng teams push for maximizing token usage, Tao's public view is to encourage verification, limit AI use for tasks that have real-life consequences and prevent low-effort papers that exhaust the review pipeline?

He also focuses on training the next generation of mathematicians and prioritizing the process of slow digestion of results, which to me sounds very.. reasonable?

Note that the SAIR lecture is from 6 months ago, this Beamer presentation is from 2 days ago.

I'm considering the possibility that it's good to break the prefix and cache because the LLM itself was rewarded (during post-training) with different prefixes/system prompts, each containing reasoning traces of the correct size.

I might be very very wrong though and LLMs disagree with me, insisting that cache is preserved and the system message doesn't have to change (even though it often contains effort level in context) if effort level changes across turns, and that all you have to do is tell the inference lib that parses think tags to early-close think tags that are too long.

Examples with inference of different reasoning effort levels is in the OpenAI docs as well - https://developers.openai.com/cookbook/articles/openai-harmo...

https://docs.vllm.ai/en/latest/features/reasoning_outputs/#a...

https://developers.openai.com/api/docs/guides/reasoning

LLM-judge/parallel branching ≠ multi-token prediction ≠ reasoning effort.

See https://developers.openai.com/cookbook/articles/openai-harmo... and src/openai/types/shared/reasoning_effort.py

The number of tokens you predict at time (multi or not) has nothing to do with whether the model wants to emit any, some or a lot of reasoning tokens in reasoning tag -- similar to how branch prediction will not really change the for loop iteration count.

Take a look at the harmony repo which specifies the internal OpenAI format - the effort level is specified in the context after the <start> tag - https://github.com/openai/harmony

Note that inference libs also have parsers that put hard limits on reasoning tokens with separate counters (similar to how you can put a limit on token generation per completion versus waiting for an <eos>).

If H800 is a memory-constrained model that NVIDIA built to avoid the Chinese export ban on H100 with equivalent fp8 performance, it makes zero sense to believe Elon Musk, Dario Armodei and Alexandr Wang's claims that DeepSeek smuggled H100s.

I feel like a fraud for telling people that DeepSeek could've smuggled chips once I took a look at the actual DeepSeek optimizations. The only reason why a team would allocate time on memory optimizations and writing NVPTX code rather than becoming much better at posttraining is if they severely struggled with memory during training.

I mean, take a look at the numbers:

https://www.fibermall.com/blog/nvidia-ai-chip.htm#A100_vs_A8...

This is a massive trick pulled by Jensen, take your best chip that's regulated by the government, find which specs are required for the regulations and then make a weaker clone that bypasses one of the specs.

You should start a blog... or maybe not - pursue the battle in academia/work and occasionally drop nuggets of wisdom like this somewhere. But do not delete them.

I updated the post with a a link to counter-argument from Sabine Hossenfelder, the arguments from Zvi and three points from my side.

I updated the post with a a link to counter-argument from Sabine Hossenfelder, the arguments from Zvi and three points from my side.

Scott worked at OpenAI Safety and he likes it: https://scottaaronson.blog/?p=8047

But is the "-ed" in worked a problem?

Thank you.

It is a question. I tried to put what my opinion is on a few statements but I absolutely cannot summarize 160 pages (Business Insider did using GPT, which I find insulting and funny) nor have a 100% opinion on something that involves national security, secrets and other stuff that I don't have access to.

You haven't read Scott's blog enough :)

He's an atheist psychiatrist. However, he enjoys how natural selection, social dynamics and reputation can also be modeled by the moral rules of most religions. For example, going to therapy isn't that different from practicing confessions in a church.