One of cancer research’s fastest growing methods is up 13x since 2021, yet fewer than 1% of its papers come from lower-income countries due to high costs.
Kalin Nonchev’s (@nonchevk) DeepSpot tackles this gap by predicting gene maps from standard tissue slides available in labs worldwide.
Kalin Nonchev’s (@nonchevk) DeepSpot tackles this gap by predicting gene maps from standard tissue slides available in labs worldwide.
15
2
72
16,650
21,000 models had to spot the moment a time series changes its behaviour, with the data arriving in real time.
ADIA Lab set the problem. 1,700 researchers worked on it on Crunch for 5 months.
The winning model scored 69.35% AUC. A random guess scores 50%.
At the @adia_lab Symposium in Abu Dhabi, Jean Herelle walked through the 10 winning methods and the top researchers behind them.
youtube.com/watch?v=lad_PK…
ADIA Lab set the problem. 1,700 researchers worked on it on Crunch for 5 months.
The winning model scored 69.35% AUC. A random guess scores 50%.
At the @adia_lab Symposium in Abu Dhabi, Jean Herelle walked through the 10 winning methods and the top researchers behind them.
youtube.com/watch?v=lad_PK…
7
2
58
9,167
AI models score higher and higher on benchmarks. GDPx tests whether they can do real specialist work, with 206 expert-level tasks across 11 professional fields.
The top frontier model scores just 12%.
SemiAnalysis published GDPx and built it with 14 specialist partners in finance and chip design. Models get up to 8 hours and access to real tools. In finance, almost every model scores zero.
The hardest tasks still need deep domain expertise and human judgment.
At Crunch, we combine both. Our researchers build the models, and we judge their work on real-world performance.
The top frontier model scores just 12%.
SemiAnalysis published GDPx and built it with 14 specialist partners in finance and chip design. Models get up to 8 hours and access to real tools. In finance, almost every model scores zero.
The hardest tasks still need deep domain expertise and human judgment.
At Crunch, we combine both. Our researchers build the models, and we judge their work on real-world performance.
11
2
50
7,038
Full report here: newsletter.semianalysis.com/p/gdpx-measuri…
0
0
14
2,704
His models did the talking.
The best model for predicting how fat cells respond to a gene switch-off came from a self-taught data scientist.
Laurent Renard is a signal processing engineer by training and an actuarial Python developer at BNP Paribas Cardif. In Part 1 of the ty ML Competition, his model finished first, more than 50% ahead of second place. He also ended 2025 at number one on a DataCrunch leaderboard, Crunch’s weekly ranking of US stocks.
In only two years he has gone from 848th to 17th of all time, collecting two gold medals and the Crunch Grandmaster title along the way.
The best model for predicting how fat cells respond to a gene switch-off came from a self-taught data scientist.
Laurent Renard is a signal processing engineer by training and an actuarial Python developer at BNP Paribas Cardif. In Part 1 of the ty ML Competition, his model finished first, more than 50% ahead of second place. He also ended 2025 at number one on a DataCrunch leaderboard, Crunch’s weekly ranking of US stocks.
In only two years he has gone from 848th to 17th of all time, collecting two gold medals and the Crunch Grandmaster title along the way.
7
3
50
8,323
AI benchmarks are going private.
60% of the new Artificial Analysis Intelligence Index is now scored on private test sets or held-out answers.
For its physics benchmark, Artificial Analysis removed the questions that no longer separated frontier models and kept the remaining new science benchmark uses research tasks that domain experts wrote and reviewed.
Crunch applies the same principle to real-world prediction problems.
If you want to know what a model can do, give it a problem it could not have trained on.
60% of the new Artificial Analysis Intelligence Index is now scored on private test sets or held-out answers.
For its physics benchmark, Artificial Analysis removed the questions that no longer separated frontier models and kept the remaining new science benchmark uses research tasks that domain experts wrote and reviewed.
Crunch applies the same principle to real-world prediction problems.
If you want to know what a model can do, give it a problem it could not have trained on.
9
3
47
7,607
Artificial Analysis Intelligence Index v5, full report: artificialanalysis.ai/articles/artif…
0
0
15
2,032
Crunch cut a tier-one global bank's currency trading costs by 4%.
700 researchers built the models, and the engine turns their forecasts into one price and a confidence score in under 60 microseconds. On a $10 billion-a-day desk, that 4% is about $10 million a year.
When the is already there.
700 researchers built the models, and the engine turns their forecasts into one price and a confidence score in under 60 microseconds. On a $10 billion-a-day desk, that 4% is about $10 million a year.
When the is already there.
48
6
277
42,781
Grateful to ADIA Lab for having us in Abu Dhabi this week.
Over five months, 1,700+ Crunchers worked on the Structural Break Challenge: Real-Time Edition. The challenge culminated at the ADIA Lab Symposium, where Jean Herelle presented the winning solutions to an audience of the field, Nobel laureates among them.
Twelve of our top Crunchers made the trip, with several presenting their own work. We’re proud of each of them and of the dedication and ingenuity that brought them there.
Over five months, 1,700+ Crunchers worked on the Structural Break Challenge: Real-Time Edition. The challenge culminated at the ADIA Lab Symposium, where Jean Herelle presented the winning solutions to an audience of the field, Nobel laureates among them.
Twelve of our top Crunchers made the trip, with several presenting their own work. We’re proud of each of them and of the dedication and ingenuity that brought them there.
Day 1 of the ADIA Lab Symposium 2026 focused on Prediction, bringing together leading researchers and experts to explore advances in data science and AI across climate science, health sciences and the digital economy.
The day also featured updates on collaborative research l Break Challenge awards and poster sessions showcasing research from the scientific community.
Watch the highlights from Day 1.
10
5
48
10,087
Horst Simon is one of the people who shaped supercomputing, the field today's AI runs on.
His forty years of work span Boeing, NASA and Berkeley Lab, and include two Gordon Bell Prizes and the TOP500 ranking of the world's fastest computers.
In Episode 2 of Minds & Machines, he t(@herelle_jean) about AI, supercomputing and the research that ADIA Lab, which he directs, has run on Crunch since 2023.
His forty years of work span Boeing, NASA and Berkeley Lab, and include two Gordon Bell Prizes and the TOP500 ranking of the world's fastest computers.
In Episode 2 of Minds & Machines, he t(@herelle_jean) about AI, supercomputing and the research that ADIA Lab, which he directs, has run on Crunch since 2023.
45
8
251
59,777
Why would a Nobel laureate co-design a competition on Crunch?
For Guido Imbens, it’s about democratizing the entire scientific process and making it less subjective.
For Guido Imbens, it’s about democratizing the entire scientific process and making it less subjective.
32
3
102
31,918
Mutian Hong is a Crunch Grandmaster, ranked 9th on the all-time leaderboard, and a PhD student at Westlake University researching large language models.
As an undergraduate at ShanghaiTech University, he founded and led DataTech, the university's data science and AI club. He ine while bringing his campus community into competitive data science. Today, ShanghaiTech ranks 4th among all institutions on Crunch.
On Crunch, he finished in the top five of two consecutive ADIA Lab challenges. In the Causal Discovery challenge, he and his teammate placed 3rd out of 1,904 participants by feeding the outputs of established causal discovery algorithms into a gradient boosting model as features. In the Structural Break challenge, they placed 5th with a ROC AUC of 89.28%. Their model is now one of the reference models in the Crunch Open Benchmark, the baseline every new participant measures themselves against.
Now at Westlake University, he brings the data intuition and empirical rigor he built through competition to his research on large language models.
As an undergraduate at ShanghaiTech University, he founded and led DataTech, the university's data science and AI club. He ine while bringing his campus community into competitive data science. Today, ShanghaiTech ranks 4th among all institutions on Crunch.
On Crunch, he finished in the top five of two consecutive ADIA Lab challenges. In the Causal Discovery challenge, he and his teammate placed 3rd out of 1,904 participants by feeding the outputs of established causal discovery algorithms into a gradient boosting model as features. In the Structural Break challenge, they placed 5th with a ROC AUC of 89.28%. Their model is now one of the reference models in the Crunch Open Benchmark, the baseline every new participant measures themselves against.
Now at Westlake University, he brings the data intuition and empirical rigor he built through competition to his research on large language models.
7
2
35
8,976