Ryu

マイコンカー(MCR)をやってました
今はLocal LLM
オープンソースは必ず勝つ

[Ryu @_ryu15_]

Anthropic が寝返ってMythos-oss 10Tを出す世界線はどこ

[Elon Musk @elonmusk]

Starlink Internet works even in the middle of nowhere

[Ryan McWhorter @rmcwhorter99]

give me six hours to chop down a tree and i will spend the first four figuring out how to give to axe to claude

[Flowers ☾ @flowersslop]

AI will get so ridiculously efficient that we will look back at GPU clusters the way we now look at these first room sized computers

[Kevin DeAnna @VDAREJamesK]

It's just illegal to be some normal guy now.

[Escha Labs @Eschalabs]

Today we are introducing Escha-W2 quantization.

A 2-bit Qwen3.6-35B-A3B model built for fast, local inference. The complete model is 12.3GB on disk—small, enough to run on a single consumer GPU — while averaging ~100% of FP8 performance across 12 benchmarks, including:

MMLU-Pro: 80.9
MATH-500: 93.8
GPQA-Diamond: 77.8
LiveCodeBench v6: 62.6
BFCL tool use: 88.9
RULER 8K–128K: 89.9
Commonsense-6: 76.1

On a single RTX 4090, the model runs:
225 tok/s single-stream generation
on 12.3GB on-disk model size

Compressing a 35B MoE model this far without collapsing its capabilities required more than a standard quantization pass.

We built an end-to-end compression system combining state-of-the-art low-bit quantization with model-aware fine tuning and recovery to preserve capabilities most vulnerable to low-bit error.

Quantizing the Qwen 3.6 35B model - from the base model to the final deployable checkpoint — took approximately 10 hours to complete.

Escha-W2 runs through a custom Qwen3-MoE runtime, which includes the weight loader, low-bit decoding kernels and serving integration required to execute the Escha format efficiently. The runtime currently supports SGLang and ZML deployment. vLLM and llama.cpp will be supported in the next release.

No retraining from scratch. No specialized accelerator. One consumer GPU.

Model download: https://t.co/hbO3VuoDGU
Runtime download: https://t.co/hrTpO4FvuA
Apache-2.0 model and runtime.

[OpenAI @OpenAI]

We’re giving scientists, mathematicians, and engineers free access to our frontier models—starting with 10,000 researchers and expanding to 100,000 through 2027.

ChatGPT for Academic Researchers is built to accelerate discovery across disciplines.

[Nic Cruz Patane @niccruzpatane]

SpaceX Starship floating in the Indian Ocean is giving Interstellar vibes.

Yes, that photo above is real.

[ADAM @adamemedia1]

It’s actually gross the amount of money China spends on things like unlimited clean energy for its citizens when it could be sending that money to Israel instead.

[Jez @JezCorden]

I think Sam Altman is patient zero with regards to AI psychosis.

[Ryu @_ryu15_]

私達はQwen3.6 27Bをもう一度作る技術を失ってしまった

[Loktar 🇺🇸 @loktar00]

Qwen 3.6 27b Bf16 vs Q4

Not cherry picked first result from each, the prompt is from @ZenMagnets check out his pinned post it's a great test!

I did throw Q4 a bone and gen twice more.. 2nd was also all white and the third was an error.

Qwen 27b has a lot of life in it still!

[OpenAI @OpenAI]

We quietly released the open-source Codex Security CLI, but Hacker News found it before we had a chance to share it here...

You can now use it to scan repositories, track findings across runs, verify fixes, and add security checks to CI/CD.

This is an early release, and we're listening to your feedback as we continue improving it.

[giulio @thelokasiffers]

I officially have llm psychosis because why does my moms profile pic look like the GPT-Live icon

[Kush @kushika_twt]

The level of marketing these days is insane.

[Tydiye @Trevor_Sande]

@BRICSinfo Just in.

USA bans all Chinese suns.

[Elon Musk @elonmusk]

Starship still floating in the ocean

[Tibo @thsottiaux]

Hello people of Sol! I've reset usage limits for all ChatGPT Work and Codex users. Together with that, a quick update on GPT-5.6 Sol usage limits.

Over the past few weeks, many of you have told us that Sol was using your Codex limits faster than expected. To be clear, we have not reduced usage on any subscription plans.

We’ve been digging into what was happening and have landed several improvements. As a result, we expect your usage to last around 18% longer during typical use of Sol. Some of you should already see significantly larger improvements from today. Tomorrow, we’ll also restore the five-hour limit that we temporarily paused while investigating.

Here’s what we found:
- GPT-5.6 Sol is much more willing to work for longer, make additional tool calls, and coordinate complex workflows across tools and subagents. That makes it better at solving hard problems, but some tasks were using far more than we intended.
- Sol also works harder at the same reasoning effort than previous models. High on Sol can use more tokens than High did on GPT-5.5.
- Programmatic tool calling, also referred to as code mode, gives Sol much more flexibility to run tool calls in parallel or continue working while waiting. But it also led to more responses per turn, more cached input tokens, and higher usage than expected.
- This was particularly noticeable when Sol was waiting for tool calls to finish or running many web searches. We’ve improved how we handle both cases and are continuing to make code mode more efficient.
- The impact was also very uneven. The median user actually found Sol quite token efficient, while some power users working on harder tasks saw their usage drain much faster. We were very focused on average and median usage before launch and missed some cases where the long tail could use significantly more usage.

Sol is a significant step forward in what Codex can do, but capability and efficiency do not always improve at the same pace, and some issues only become clear once people are using the model at real-world scale. We should have recognized this sooner and been more upfront about it.

You keep pushing the frontier and we’ll keep improving efficiency and sharing updates as we go.

[Elon Musk @elonmusk]

We’re sending a ship out to recover Starship

[Charles Hoskinson @IOHK_Charles]

American healthcare in a single sentence

[Polymarket @Polymarket]

JUST IN: China’s Moonshot AI reportedly seeks more advanced Nvidia chips to train Kimi K4, despite U.S. export restrictions.

[Elon Musk @elonmusk]

Interesting.

Grok 4.6 releases around August 7. This will be the 1.5T model with significantly improved SFT & RL.

Grok 4.7 will be the 2.1T model released a few weeks later. This will be better than 4.6 in every way, except slightly slower to serve, albeit with even better token efficiency.