arXiv now allows researchers to submit code with their manuscripts

Posted on

Papers with Code today announced that preprint paper archive arXiv will now allow researchers to submit code alongside research papers, giving computer scientists an easy way to analyze, scrutinize, or reproduce claims of state-of-the-art AI or novel advances in what’s possible.

An assessment of the AI industry released a week ago found that only 15% of papers submitted by researchers today publish their code.

Maintained by Cornell University, arXiv hosts manuscripts from fields like biology, mathematics, and physics, and it has become one of the most popular places online for artificial intelligence researchers to publicly share their work. Preprint repositories give researchers a way to share their work immediately, before undergoing what can be a long peer review process as practiced by reputable scholarly journals. Code shared on arXiv will be submitted through Papers with Code and can be found in a Code tab for each paper.

“Having code on arXiv makes it much easier for researchers and practitioners to build on the latest machine learning research,” Papers with Code cocreator Robert Stojnic said a blog post today. “We also hope this change has ripple effects on broader computational science beyond machine learning. Science is cumulative. Open science, including making available key artefacts such as code, helps to accelerate progress by making research easier to build upon.”

Started in 2018, Papers with Code focuses on encouraging reproducibility of AI model results and, as the name states, submitting research with code. The Papers with Code website shares nearly 2,000 papers and code from across major fields in AI like natural language processing, computer vision, adversarial machine learning, and robotics. Papers with Code was initially founded in part by members of Facebook AI Research. Last year, Facebook and Papers with Code launched PyTorch Hub to encourage reproducibility.

In the past year or


AI researchers challenge a robot to ride a skateboard in simulation

Posted on

AI researchers say they’ve created a framework for controlling four-legged robots that promises better energy efficiency and adaptability than more traditional model-based gait control of robotic legs. To demonstrate the robust nature of the framework that adjusts to conditions in real time, AI researchers made the system slip on frictionless surfaces to mimic a banana peel, ride a skateboard, and climb on a bridge while walking on a treadmill. An Nvidia spokesperson told VentureBeat that only the frictionless surface test was conducted in real life because of limits placed on office staff size due to COVID-19. The spokesperson said all other challenges took place in simulation. (Simulations are often used as training data for robotics systems before those systems are used in real life.)

“Our framework learns a controller that can adapt to challenging environmental changes on the fly, including novel scenarios not seen during training. The learned controller is up to 85% more energy-efficient and is more robust compared to baseline methods,” the paper reads. “At inference time, the high-level controller needs only evaluate a small multi-layer neural network, avoiding the use of an expensive model predictive control (MPC) strategy that might otherwise be required to optimize for long-term performance.”

The quadruped model is trained in simulation using a split-belt treadmill with two tracks that can change speed independently. That training in simulation is then transferred to a Laikago robot in the real world. Nvidia released video of simulations and laboratory work Monday, when it also unveiled AI-powered videoconferencing service Maxine and the Omniverse simulated environment for engineers in beta.

A paper detailing the framework for controlling quadruped legs was published a week ago on preprint repository arXiv. AI researchers from Nvidia; Caltech; University of Texas, Austin; and the Vector Institute at the University of Toronto contributed to the


Newly Discovered ‘Extreme’ Alien Planet Is Super Hot At 5,800 Fahrenheit, Researchers Reveal

Posted on


  • CHEOPS has released the results of its observation on alien planet WASP-189b
  • WASP-189b’s orbit is tilted dramatically and orbits its star every 2.7 Earth days
  • WASP-189b has temperatures reaching 5,800 Fahrenheit

The European Space Agency’s Characterizing Exoplanet Satellite (CHEOPS) has recently discovered an alien planet about 1.6 times the size of Jupiter. Aside from having a strange orbit, it is also scorching hot.

WASP-189b, the newly discovered alien planet, was first detected in 2018 and has been recorded to have temperatures reaching 5,800 Fahrenheit — almost as hot as Earth’s outer core and is even hot enough to turn iron into gas, ESA’s study revealed.

Aside from having a size comparable to Jupiter, the exoplanet is also considered a “Hot Jupiter” due to its extremely short orbital period (2.7 Earth days). A Hot Jupiter is a gas planet with a “Jupiter-like” size that orbits very close to its star.

The star which the alien planet orbits is super hot — more than 2000 degrees hotter than the Sun — so it carries a bluish hue. CHEOPS’ observations show that it is not perfectly round and is larger and cooler at its equator than at its poles. 

“Only a handful of planets are known to exist around stars this hot, and this system is by far the brightest,” says Monika Lendl, an astrophysicist at the University of Geneva in Switzerland.

“WASP-189b is also the brightest hot Jupiter that we can observe as it passes in front of or behind its star, making the whole system really intriguing.”

Unlike Earth’s solar system, where planets orbit at the sun’s equator, WASP-189b orbits its star in such a dramatic tilt that it brings it closer to the star’s poles. This characteristic makes scientists suspect that WASP-189b formed somewhere far away from the


There’s a giant ‘Green Banana’ off Florida’s coast, and researchers have finally gotten to the bottom of it

Posted on

Credit: CC0 Public Domain

If you haven’t heard of the “Green Banana blue hole” you might imagine a tropical cocktail you can order in Key West, or a dessert you ordered after a night on Bourbon Street.

Forget that. This Green Banana is actually a mysterious sink hole. More specifically, it’s a huge, underwater cavern off the coast of Florida that humans had never fully explored—until last month.

Scientists say the Green Banana could hold clues to the formation of toxic red tides, algae blooms that are devastating to Florida’s shoreline, and the extent of the aquifer that supplies the state with most of its drinking water.

Maybe even the origins of life.

Blue holes—sink holes that form under water—are not unusual in the Gulf of Mexico. In the mid-1970s, a boat captain sailing about 60 miles west of Sarasota spotted one about 160 feet under water, and an unripe banana peel floating above it. It became known as the Green Banana.

Scientists believe it may have formed more than 10,000 years ago when a sink hole opened to form a cavern 265 feet deep and 425 feet below the surface of the Gulf, further than typical scuba divers are capable of reaching.

It’s not just the depth of the Green Banana that’s a challenge for explorers. It’s wide base created by an hourglass shape had never been fully explored until advanced diver Marty Watson did it in August with a team of scientists and researchers.

“What’s it like?” Watson asked. “I’m not an astronaut, but it’s got to be the closest thing in the world next to it.”

Blue holes are thought to be ecological hot spots whose nutrients help supply the food chain around the world. It starts with the phytoplankton that feed on those nutrients, which attracts