Learning concepts with energy functions
📰 OpenAI News
OpenAI develops an energy-based model that can learn to identify and generate concepts with few demonstrations
Action Steps
- Understand the concept of energy-based models
- Explore the application of energy functions in learning concepts
- Analyze the results of the model in learning and generating concepts
- Investigate cross-domain transfer of learned concepts
Who Needs to Know This
AI researchers and engineers can benefit from this concept learning technique to improve agent understanding and reasoning, while product managers can explore applications in various domains
Key Insight
💡 Energy-based models can quickly learn to identify and generate instances of concepts, enabling cross-domain transfer and improved agent understanding
Share This
💡 Energy-based models can learn concepts with few demos! #AI #ML
Key Takeaways
OpenAI develops an energy-based model that can learn to identify and generate concepts with few demonstrations
Full Article
# Learning concepts with energy functions | OpenAI
[Skip to main content](https://openai.com/index/learning-concepts-with-energy-functions#main)
[](https://openai.com/)
* [Research](https://openai.com/research/index/)
* Products
* [Business](https://openai.com/business/)
* [Developers](https://openai.com/api/)
* [Company](https://openai.com/about/)
* [Foundation(opens in a new window)](https://openaifoundation.org/)
Log in[Try ChatGPT(opens in a new window)](https://chatgpt.com/)
* Research
* Products
* Business
* Developers
* Company
* [Foundation(opens in a new window)](https://openaifoundation.org/)
[Try ChatGPT(opens in a new window)](https://chatgpt.com/)Login
OpenAI
Table of contents
* [How it works](https://openai.com/index/learning-concepts-with-energy-functions#how-it-works)
* [Single network training](https://openai.com/index/learning-concepts-with-energy-functions#single-network-training)
* [Key results](https://openai.com/index/learning-concepts-with-energy-functions#key-results)
* [Next steps](https://openai.com/index/learning-concepts-with-energy-functions#next-steps)
November 7, 2018
[Publication](https://openai.com/research/index/publication/)
# Learning concepts with energy functions
[Read paper(opens in a new window)](https://arxiv.org/abs/1811.02486)[View code(opens in a new window)](https://sites.google.com/site/energyconceptmodels/)
Loading…
Share
We’ve developed an[energy-based model(opens in a new window)](https://arxiv.org/abs/1708.06008)that can quickly learn to identify and generate instances of concepts, such as near, above, between, closest, and furthest, expressed as sets of 2d points. Our model learns these concepts after only five demonstrations. We also show cross-domain transfer: we use concepts learned in a 2d particle environment to solve tasks on a 3-dimensional physics-based robot.
Many hallmarks of human intelligence, such as generalizing from limited experience, abstract reasoning and planning, analogical reasoning, creative problem solving, and capacity for language require the ability to consolidate experience into _concepts_, which act as basic building blocks of understanding and reasoning. Our technique enables agents to learn and extract concepts from tasks, then use these concepts to solve other tasks in various domains. For example, our model can use concepts learned in a two-dimensional particle environment to let it carry out the same task on a three-dimensional physics-based robotic environment—**without retraining in the new environment.**
Loading...
This work uses energy functions to let our agents learn to _classify_ and _generate_ simple concepts, which they can use to solve tasks like navigating between two points in dissimilar environments. Examples of concepts include visual (“red” or “square”), spatial (“inside”, “on top of”), temporal (“slow”, “after”), social (“aggressive”, “helpful”) among others. These concepts, once learned, act as basic building blocks of agent’s understanding and reasoning, as shown in other research from[DeepMind(opens in a new window)](https://deepmind.com/blog/imagine-creating-new-visual-concepts-recombining-familiar-ones/)and[Vicarious(opens in a new window)](https://www.vicarious.com/2018/02/07/learning-concepts-through-sensorimotor-interactions/).
[](https://openai.com/)
* [Research](https://openai.com/research/index/)
* Products
* [Business](https://openai.com/business/)
* [Developers](https://openai.com/api/)
* [Company](https://openai.com/about/)
* [Foundation(opens in a new window)](https://openaifoundation.org/)
Log in[Try ChatGPT(opens in a new window)](https://chatgpt.com/)
* Research
* Products
* Business
* Developers
* Company
* [Foundation(opens in a new window)](https://openaifoundation.org/)
[Try ChatGPT(opens in a new window)](https://chatgpt.com/)Login
OpenAI
Table of contents
* [How it works](https://openai.com/index/learning-concepts-with-energy-functions#how-it-works)
* [Single network training](https://openai.com/index/learning-concepts-with-energy-functions#single-network-training)
* [Key results](https://openai.com/index/learning-concepts-with-energy-functions#key-results)
* [Next steps](https://openai.com/index/learning-concepts-with-energy-functions#next-steps)
November 7, 2018
[Publication](https://openai.com/research/index/publication/)
# Learning concepts with energy functions
[Read paper(opens in a new window)](https://arxiv.org/abs/1811.02486)[View code(opens in a new window)](https://sites.google.com/site/energyconceptmodels/)
Loading…
Share
We’ve developed an[energy-based model(opens in a new window)](https://arxiv.org/abs/1708.06008)that can quickly learn to identify and generate instances of concepts, such as near, above, between, closest, and furthest, expressed as sets of 2d points. Our model learns these concepts after only five demonstrations. We also show cross-domain transfer: we use concepts learned in a 2d particle environment to solve tasks on a 3-dimensional physics-based robot.
Many hallmarks of human intelligence, such as generalizing from limited experience, abstract reasoning and planning, analogical reasoning, creative problem solving, and capacity for language require the ability to consolidate experience into _concepts_, which act as basic building blocks of understanding and reasoning. Our technique enables agents to learn and extract concepts from tasks, then use these concepts to solve other tasks in various domains. For example, our model can use concepts learned in a two-dimensional particle environment to let it carry out the same task on a three-dimensional physics-based robotic environment—**without retraining in the new environment.**
Loading...
This work uses energy functions to let our agents learn to _classify_ and _generate_ simple concepts, which they can use to solve tasks like navigating between two points in dissimilar environments. Examples of concepts include visual (“red” or “square”), spatial (“inside”, “on top of”), temporal (“slow”, “after”), social (“aggressive”, “helpful”) among others. These concepts, once learned, act as basic building blocks of agent’s understanding and reasoning, as shown in other research from[DeepMind(opens in a new window)](https://deepmind.com/blog/imagine-creating-new-visual-concepts-recombining-familiar-ones/)and[Vicarious(opens in a new window)](https://www.vicarious.com/2018/02/07/learning-concepts-through-sensorimotor-interactions/).
DeepCamp AI