Demos

Generate climate policies using Cognizant NeuroAI integrated with the En-ROADS simulator for informed decision-making.

Optimize land-use policies to balance carbon impact and land-use changes for specific locations.

Explore AI-generated NPI strategies for regions worldwide, balancing case reduction and intervention stringency over time.

See how ESP outperforms direct evolution and PPO, finding better solutions with fewer samples.

See how ESP excels in the Flappy Bird domain, using a surrogate to regularize learning and generalize behaviors for optimal strategies.

Explore the technical details of the competition, including evaluation methods, results, and interactive leaderboards from different phases.

Optimize crop growth with Evolutionary AI to enhance yield, flavor, and efficiency in controlled environments.

Optimize webpage designs using evolutionary AI to uncover creative, high-performing solutions that maximize conversions.

Draw a character and see how multitask and single-task models interpret it across 20 alphabets, revealing insights into character recognition and model learning.

Explore how the CMSR approach leverages coevolutionary methods to develop modular deep learning architectures for improved multitask learning.

See how the CTR approach evolves custom network designs using shared modules to enhance multitask learning.

See how the composite novelty method explores useful trade-offs and diverse behaviors to uncover better solutions in complex optimization problems.

Watch how the composite novelty method outperforms standard optimization by overcoming deception to discover optimal 8-line sorting networks.

Compose a melody and let an evolved LSTM network build on it, generating a unique musical sequence inspired by your input.

See how evolutionary methods uncover more powerful LSTM architectures, improving performance in language modeling tasks.

Watch how evolution explores a search space to uncover effective solutions, visualized through an animated search for optimal webpage designs.

See how evolutionary algorithms accumulate confidence over generations, improving reliability in noisy and uncertain environments.

Explore how asynchronous multi-arm bandit algorithms improve candidate evaluations by allocating resources efficiently, boosting performance and reliability.

See how the asynchronous multi-arm bandit algorithm maximizes overall conversion rates during campaigns by efficiently allocating resources across generations.

Explore how Best Arm Identification Mode enhances reliability by refining top candidates, ensuring the best solution is chosen with confidence.

Watch how evolution uncovers optimal trade-offs between objectives, improving solution quality over time in the COVID-19 Dashboard demo.