Generalizing from simulation

OpenAI Blog2017-10-19 15:000 views

Our latest robotics techniques allow robot controllers, trained entirely in simulation and deployed on physical robots, to react to unplanned changes in the environment as they solve simple tasks. That is, we’ve used these techniques to build closed-loop systems rather than open-loop ones as before.

Source：https://openai.com/index/generalizing-from-simulation

My favorites

PreviousLearning a hierarchy NextSim-to-real transfer of robotic control with dynamics randomization

One sentence can flip your AI's decision. Here is how we catch it.

Decision Steering Gate v2, plus IFlippedJev and NotClaim: a local claim detector and two open benchmarks for decision st

0dev.to Security

Building a Secret Santa draw in Next.js: derangements, HMAC links, and not leaking emails

Every December someone ships a Secret Santa app, and every December the same three things go wrong: the shuffle loops fo

0dev.to Security

Scooter's Coffee Menu (2026): Drinks, Prices, Calories and Best Orders

The Scooter's Coffee menu is big. It offers espresso, cold brew, smoothies, drinks with Red Bull® Infusions, teas, lemon

0dev.to Python

Circular imports in Python: why every big Python repo has them, and the few that actually break

TL;DR All 51 popular Python repos I checked (1,000+ GitHub stars each) have circular import chains between files; in the

0dev.to Python

cf-cache-status DYNAMIC: Why Cloudflare Isn't Caching Your HTML (and How to Fix It)

If you run curl -I against a site behind Cloudflare and see cf-cache-status: DYNAMIC , Cloudflare isn't caching that res

0dev.to DevOps

VoIP One-Way Audio and Calls That Drop at 32 Seconds: What Causes Them and How to Fix Them

VoIP one-way audio and calls that drop at about 32 seconds are nearly always NAT problems, not a bad internet connection