Wang Liang

nlper , phd@hust, tencent,
github:

[Patrick Kidger @PatrickKidger]

I've had [what feels like] approximately every AI+bio company ask me for any talent referrals recently. It's a seller's market right now.
If you have at least a small portfolio (strong PhD or a couple years in industry) + are willing to work at least a 9-7 then you're golden.
1/2

[Wang Liang @maris205]

I just published Hands-On Embodied AI for Drones: Powered by AirSim & IONET https://t.co/ZKSj9Eroro

[Lisan al Gaib @scaling01]

A 27B model using a mere 3.9GBs

[小互 @xiaohu]

他妈的，天塌了

GPT 这么快就被污染了

龟儿子把Token翻译成词元了

草他妈的...

[Adib @adibvafa]

Exactly what the field needed!

@TrillionLabs brought all scattered biological databases and instruction datasets together for a unified training corpus

[Wang Liang @maris205]

no no no, 是放高利贷的，法人对赌，5-12%的年化利率，哈哈哈

[墙国反贼 @Qiangguofanzei]

小红书上描绘中国投资人，笑死我了，非常准确

[Johannes Tscharn @JohannesTscharn]

Augmented Reality is the natural interface for robotics.

Not only for giving commands, but also to visualise hardware data (like LiDAR) natively in 3D.

@specs AR glasses, @UnitreeRobotics Go2 Lite quadruped, @dimensionalos OS for the robot

[Federico De Ponte @fedeponte1]

OpenPaper is Claude Code for research papers.

One question in. Five minutes later: a complete 8,330-word research paper with 30 citations, every one a real published paper with a working DOI, checked against 500M+ academic sources. Zero hallucinated references.

18 agents run the research in parallel. Open source.

https://t.co/bqfWZDXC9x

[Wang Liang @maris205]

gpt-5.6-sol ultra的数学能力太离谱了，根据之前的方向，已经写了13个新论文了。。。。。

[Derya Unutmaz, MD @DeryaTR_]

The moment of truth in Math AI: GPT-5.6 Sol Ultra solved a 50-year-old math problem with a proof!

This is also the end of the era for AI critics who foolishly claimed AI can never make discoveries, be innovative, or solve the hardest problems. Years of coping battle is now over.

[Wang Liang @maris205]

GPT-5.6 明显强了不少，思路清晰很多

[Mile Sikic @msikic]

@Anthropic is betting that biology is AI's next frontier. I agree—but there are two streams: building virtual bioinformaticians that combine existing knowledge & tools, or AI that discovers new biology in close collaboration with wet-lab scientists. Our focus is the latter.

We're hiring interns, research officers & postdocs at the A*STAR Genome Institute of Singapore. Projects:
• RNA therapeutics—optimizing mRNA for drugs & vaccines
• RNA drugs—structure prediction & small-molecule binding
• World models for biology
• Next-gen read error correction (cancer genomes) / HERRO v2

Looking for candidates with a proven AI track record—papers at top AI conferences and in journals. If that's you, get in touch. Please share!! 🔁

[Kexin Huang @KexinHuang5]

Today, we're excited to share that Biomni is published in @ScienceMagazine.

Biomedical research is still fragmented, manual, and difficult to scale. In this work, we introduce Biomni - the first general-purpose biomedical AI agent with an integrated biology environment that can reason, plan, and execute end-to-end scientific workflows.

We show that, with the right environment and harness, AI can automate large-scale omics analyses, orchestrate laboratory robotics, optimize molecular properties, and even train new AI models for biology.

We also introduce a reinforcement learning recipe for continually improving biomedical AI agents, enabling open-source models to achieve frontier-level performance.

It's surreal to look back. We started the Biomni project in early 2024, when agentic AI was still nascent. It is exciting to see tens of thousands of biologists collaborating with agents every day to accelerate science.

Try Biomni: https://t.co/Pzm8TVlkNA
Read more: https://t.co/pRstxzPKDi

This work is not possible without this truly inter-disciplinary team: @serena2z @hcwww_ @YuanhaoQ Minta Lu, Ryan Li, @yusufroohani Lin Qiu @shiyi_c98 Gavin Junze Di @rickwierenga @kavi_deniz Sherry @TianweiShe Shruti Jennefer Xin Zhou @MWheelerMD Jon Bernstein @MengdiWang10 @PengHeAtlas @zhou_jingtian @SnyderShot @lecong Aviv Regev @jure

@StanfordAILab @genentech @phylo_bio @arcinstitute @UW @berkeley_ai @RetroBio_ @tamarindbio @Princeton @UCSF

[Synthetic Sciences (YC W26) @SynScience]

Introducing OpenScience. A better, open-source Claude Science.

• Any model: GLM, Kimi, DeepSeek, Claude, GPT, your own fine-tune. Switching is one flag.
• 250+ research skills across ML, comp bio, cheminformatics. All readable, editable, extensible.
• No throttling, no gatekeeping, no one vendor deciding what science is okay.
• Native Atlas integration: many agents, one shared reproducible research graph.
• Runs on your infra. Your data stays yours.

Scientific AI should be open. One company shouldn't own the tools the rest of us discover with, or decide who gets to.

[Jérôme Lecoq @LecoqJerome]

I tried this out quite extensively over the course of multiple weeks before their official launch. I had agents running for multiple days uninterrupted in the end. I will share more what I learned soon. How we do science is changing very rapidly.

[Georgia Channing @cgeorgiaw]

Seems like no one's noticed the 80TB of astrophysics data from 30+ sources that just dropped on @huggingface.

...and you only need ~4GB of RAM to load it.

We're talking over 80TB of galaxy imagery taken across the spectrum, spectra of galaxies and stars, time series of variable stars, and a whole zoo of assorted measurements and physical data.

And all of it can now be wrangled on your laptop, thanks to Multimodal Universe's just released cross-matching. SDSS x Gaia means you can match 800k objects against 122M objects, and it never climbs above ~4GB of RAM.

Huge congrats to @smith42mike for leading this and making the world of astro accessible to probably 10,000x more people. Let's discover some shit

[Jeremy @jeremyli__]

Glad to get this work w/ @andrewho03 out – GeneBench-Pro is an updated, verified version of our benchmark measuring performance of frontier models on difficult long-horizon tasks in comp bio. Link to personal blog with some thoughts below.

https://t.co/sDugQVtsEU

[Claude @claudeai]

Introducing Claude Science, a new app designed with every stage of research in mind.

Artifacts traced to their code, environments managed on demand, and 60+ optional scientific databases that you can connect.

Available now in beta.

[Gina El Nesr @ginaelnesr]

Dyna-1 was upgraded to an MIT license as a result 😍