We use reinforcement learning (RL) to learn dexterous in-hand manipulation policies that can perform vision-based object reorientation on a physical Shadow Dexterous Hand. The training is performed in a simulated environment in which we randomize many of the physical properties of the system such as friction coefficients and an object’s appearance. Our policies transfer to the physical robot despite being trained entirely in simulation. Our method does not rely on any human demonstrations, but many behaviors found in human manipulation emerge naturally, including finger gaiting, multi-finger coordination, and the controlled use of gravity. Our results were obtained using the same distributed RL system that was used to train OpenAI Five. We also include a video of our results: https://youtu.be/jwSbzNHGflM .

We introduce Codex, a GPT language model finetuned on publicly available code from GitHub, and study its Python code-writing capabilities. A distinct production version of Codex powers GitHub Copilot. On HumanEval, a new evaluation set we release to measure functional correctness for synthesizing programs from docstrings, our model solves 28.8% of the problems, while GPT-3 solves 0% and GPT-J solves 11.4%. Furthermore, we find that repeated sampling from the model is a surprisingly effective strategy for producing working solutions to difficult prompts. Using this method, we solve 70.2% of our problems with 100 samples per problem. Careful investigation of our model reveals its limitations, including difficulty with docstrings describing long chains of operations and with binding operations to variables. Finally, we discuss the potential broader impacts of deploying powerful code generation technologies, covering safety, security, and economics.

Sleep spindles are discrete, intermittent patterns of brain activity that arise as a result of interactions of several circuits in the brain. Increasingly, these oscillations are of biological and clinical interest because of their role in development, learning, and neurological disorders. We used an internet interface to ‘crowdsource’ spindle identification from human experts and non-experts, and compared performance with 6 automated detection algorithms in middle-to-older aged subjects from the general population. We also developed a method for forming group consensus, and refined methods of evaluating the performance of event detectors in physiological data such as polysomnography. Compared to the gold standard, the highest performance was by individual experts and the non-expert group consensus, followed by automated spindle detectors. Crowdsourcing the scoring of sleep data is an efficient method to collect large datasets, even for difficult tasks such as spindle identification. Further refinements to automated sleep spindle algorithms are needed for middle-to-older aged subjects.

We present a fast and accurate algorithm for computing the 2D pose of objects in images called cascaded pose regression (CPR). CPR progressively refines a loosely specified initial guess, where each refinement is carried out by a different regressor. Each regressor performs simple image measurements that are dependent on the output of the previous regressors; the entire system is automatically learned from human annotated training examples. CPR is not restricted to rigid transformations: 'pose' is any parameterized variation of the object's appearance such as the degrees of freedom of deformable and articulated objects. We compare CPR against both standard regression techniques and human performance (computed from redundant human annotations). Experiments on three diverse datasets (mice, faces, fish) suggest CPR is fast (2-3ms per pose estimate), accurate (approaching human performance), and easy to train from small amounts of labeled data.

Making language models bigger does not inherently make them better at following a user's intent. For example, large language models can generate outputs that are untruthful, toxic, or simply not helpful to the user. In other words, these models are not aligned with their users. In this paper, we show an avenue for aligning language models with user intent on a wide range of tasks by fine-tuning with human feedback. Starting with a set of labeler-written prompts and prompts submitted through the OpenAI API, we collect a dataset of labeler demonstrations of the desired model behavior, which we use to fine-tune GPT-3 using supervised learning. We then collect a dataset of rankings of model outputs, which we use to further fine-tune this supervised model using reinforcement learning from human feedback. We call the resulting models InstructGPT. In human evaluations on our prompt distribution, outputs from the 1.3B parameter InstructGPT model are preferred to outputs from the 175B GPT-3, despite having 100x fewer parameters. Moreover, InstructGPT models show improvements in truthfulness and reductions in toxic output generation while having minimal performance regressions on public NLP datasets. Even though InstructGPT still makes simple mistakes, our results show that fine-tuning with human feedback is a promising direction for aligning language models with human intent.