Researchers stretch LeCun’s JEPA AI into a universal world model that works from physics to biology

Researchers have expanded Yann LeCun’s JEPA architecture into a system that functions across seven distinct fields, from physics to biology, while identifying a liver cancer treatment candidate that outperformed existing methods in lab tests.

Breaking the single-prediction bottleneck

Standard world models typically funnel all data into one abstract prediction. This approach often allows simple patterns to drown out complex ones. The new system, JEPA-Anything, changes this by splitting the predicted state into several parts. Each part is handled by its own prediction module.

Added constraints force these modules to capture different aspects of the data rather than learning the same information repeatedly. The model then reassembles these partial predictions into a complete picture. The team does not assign specific meanings to these parts beforehand; the roles emerge naturally during training. They only adjust how the data is prepared for each specific field.

Performance gains across physics, biology, and robotics

The team compared JEPA-Anything against a standard JEPA model using identical architecture, data, and training conditions. Dynamic systems showed the clearest improvements. In a simplified Pong environment with targeted interventions, prediction error dropped by 35 percent. For combinations of interventions the model had never seen during training, the error fell by 13 percent.

The system consistently beat the baseline across ten test tasks spanning physics, robotics, and weather forecasting. On the Burgers equation, a common fluid dynamics benchmark, error fell by nearly half in a separate evaluation. The advantage held for over 50 prediction steps before shrinking to about three percent. The method also scored best in simulations of water, quartz, acetaminophen, and benzene, even after 100 steps.

For single-cell data, the model assigned cell types more reliably. On clinical data, it predicted more than 1,000 possible disease events slightly better than the baseline. The difference on image tasks was small. In locomotion planning for simulated walking robots, JEPA-Anything won in two of three environments, while the standard model came out ahead in the third.

A liver cancer candidate and orbital physics

The team’s boldest claim comes from liver cancer research. Researchers analysed partial predictions the model had learned from biological data including gene activity, protein levels, and CRISPR screens. The top candidate paired IL-18, a signalling molecule that activates immune cells, with blockade of the enzyme CD73, which tumors use to suppress nearby immune responses.

The team tested the combination on liver cancer cells co-cultured with immune cells, on organoids and tumor tissue from three patients each, and in mice. In the organoids and tissue samples, the combination killed more tumor cells than either IL-18 or CD73 blockade alone. T cells and natural killer cells showed stronger activation. The study does not establish whether this could become an actual therapy.

In a second case, the researchers trained the model on simulated orbits without giving it any physical quantities. The learned patterns turned out to be a near-exact match for Kepler’s third law, which states bodies on larger orbits move much more slowly. The law sets orbital frequency at orbit size to the power of minus 1.5; the model landed on minus 1.4991. The team only evaluated one training run and picked the one with the lowest error.

What it means

The authors caution that a clean separation of learned parts does not mean they capture real cause-and-effect relationships. It remains an open question when such a model becomes reliable enough to guide experiment design. That is the team’s long-term goal: AI agents would use JEPA-Anything to propose and rank experiments, then feed results back into the model. Code and models are publicly available.

LeCun proposed JEPA in 2022 as an alternative to generative models. In June 2025, Meta released V-JEPA 2, a JEPA video model with 1.2 billion parameters that controlled robotic arms in unfamiliar environments without additional training. In November 2025, LeCun and Randall Balestriero followed up with LeJEPA, a theoretical foundation designed to keep training stable without the usual workarounds. LeCun is now pursuing the approach through his startup AMI Labs, which raised over a billion dollars in March 2026 to build world models.

Google Deepmind had already tested an AI-generated cancer hypothesis in the lab in October 2025. Its Gemma-based model C2S-Scale 27B proposed the drug silmitasertib to make tumor cells more visible to the immune system, and experiments with human cell models confirmed the prediction. Google Deepmind’s multi-agent system Co-Scientist now plans experiments and operates lab equipment, covering parts of the loop the JEPA-Anything team wants to build. Loading samples into the machines, though, still requires humans.