Bo Wang

This account is verified. Learn more

@BoWang87

Prof @UofT | Co-Founder & Chief AI Scientist @Xaira_Thera | Building first Virtual Cell | AI & Bio & Healthcare | Co-Inventor of ScGPT, MedSAM, BIOREASON

San Francisco, CA Joined September 2016

3.3K Following 32K Followers 2.2K Posts

Latest posts 5

Pinned post

Bo Wang @BoWang87 Oct 23

The biggest bottleneck in AI for biology is still the wet lab.

AI can generate more ideas than our labs can test. Closing that gap is how we turn better models into faster discovery.

Today, we’re introducing X-Automate at Xaira: bringing agentic AI and physical robotics into the lab. 🧵

2:47

Bo Wang @BoWang87 2d

AI reviewer for AI paper 😅

Arth Singh 🦉 Verified account @iarthsingh 2d

Guys this ICLR paper got a 4 ? WTF ??? Generational AI slop it was

View quoted post on X

Bo Wang @BoWang87 6d

GPT-7 will be built by GPT-6x for sure! 😅

Rikiya Takehi @rikiyatakehi 6d

We asked GPT-6 Astra to edit Qwen3.5-0.8B weights to improve math.
It directly changed only 0.016% of weights, fixing 0.8B's reasoning loops & improving math up to 6x and even instruction following 3x, without gradients or training.

We call this Weight-Space Programming: frontier LLMs can now natively read/write neural network weights! 🧵
Blog: rikiyat.github.io/blog/weight-sp…

View quoted post on X

Bo Wang @BoWang87 Oct 23

Crazy! Kimi-3.1 Pro 🔥

Kimi.ai @Kimi_Moonshot Oct 23

Meet Kimi K3.1 Pro, our most capable 3T-class model.

View quoted post on X

Bo Wang @BoWang87 Oct 23

We connected X-Scientist to our lab robots and watched the lab come alive. 🦾🧪

AI coordinating workflows. Robots moving samples. Experiments running into the night.

It’s incredibly rewarding to watch the future of biotech unfold right in front of us.

2:18

Bo Wang Verified account @BoWang87 Oct 21

Meet X-Scientist: our step toward a self-improving virtual drug hunter!

Today, we’re introducing Xaira’s agentic AI platform for drug discovery:

🧠 Beats Codex using the same Astra model by 7.4 percentage points on DrugDiscoveryBench.

🧬 Nominates targets ~10× faster than human curation, with a comparable wet-lab hit rate.

🔁 Conducts autonomous research, outperforming AutoResearch and AutoScientists in the PerturbQA benchmark.

Here’s what we built, what we tested, and where we’re going. 🧵

View quoted post on X