Every system has blind spots.
Places where the expected breaks. Where signals turn up from noise. Where models, built in silence, speak louder than forecasts.
CrunchDAO exists in those spaces between what’s measurable and what matters.
This week, a few of those signals surfaced.
Here’s what went down ↓
Autoimmune Disease Machine Learning Crunch #2: OOS Leaderboard Released
The Out-of-Sample leaderboard for the Autoimmune Disease ML Challenge #2 is now live.
Participants predicted expression for 2,000+ genes using spatial transcriptomics and scRNA-seq data, tackling one of the hardest problems in autoimmune science.
Datacrunch Equity Market Neutral Leaderboard Update
Top performers this week made significant gains across all targets.
Standouts include etothexipi, yellow-filip, selective-zhu, and amused-david.
Plus, March payouts are live
Crunch at Solana Incubator
CrunchDAO joined the Solana Incubator Demo Day where our CSO, Peter Cotton, showcased the platform and what we’ve been building over the last few months as part of Cohort 2 of the Solana Incubator program.
CrunchDAO x Arcium X Space: Making Private ML Actually Work
A live session with Arcium unpacked how encrypted compute and MPC frameworks turn CrunchDAO’s privacy-first vision into reality. Enabling secure, verifiable ML at scale.
→ Replay
Autoimmune Disease Broad 1 Leaderboard Spotlight: Team PathBio
Team PathBio brought together researchers from Stanford, Columbia, Zhejiang, and more. Building a spatial transcriptomics foundation model from standard histology slides.
Their work focused on predicting gene expression at the single-cell level, combining vision, computational pathology, and spatial biology.
Led by:
Sen Yang, Senior Research Professional at Stanford, expert in biomedical imaging & computational pathology
Team Members:
• Jinxi Xiang — Computer vision & medical imaging
• Wei Yuan — Computational pathology
• Yijiang Chen — Spatial biology & oncology
• Xiyue Wang — Weak supervision in pathology
A cross-disciplinary effort pushing the frontiers of autoimmune disease research.
Autoimmune Disease Broad 1 Leaderboard Spotlight: Manfred Seiwald
Manfred Seiwald transitioned from decades of enterprise software at Skidata AG to biomedical ML at Fortelny Lab, University of Salzburg.
In Challenge #1, he used a ResNet-50 model pretrained on histology patches to predict the expression of 460 genes from colon tissue — no direct measurements, just image-driven inference.
His sharp, effective approach placed him on the Broad #1 leaderboard, proving that experience and focus still move the needle in complex research.
Peter Cotton at QuantVision 2025
Dr. Peter Cotton, Chief Scientific Officer at CrunchDAO, is set to engage in a thought-provoking debate with Dr. Igor Halperin at QuantVision 2025, hosted by Fordham University. The session, titled “Financial Mathematics or Financial Physics?”, will explore the application of machine learning in quantitative finance.
The debate is scheduled for March 27, 2025, at QuantVision 2025, Fordham’s premier quantitative conference.
Crunch is Hiring
Current open roles:
- Quant Developer
- Business Development Manager
- Developer Relations Lead
- Executive Assistant
- Product Manager
Up Next
→ CrunchDAO will be at Token2049 Dubai, stay tuned for updates from the ground.
→ More spotlights from recent challenges will roll out over the coming weeks.
→ ADIA Lab challenge payouts are also on the way.
For questions, feedback, or just to jam on models —
Join the conversation on Discord.