InterACT
Welcome to the Interactive Agents and Collaborative Technologies Laboratory at UC Berkeley! Our goal is to enable AI agents to work with, around, and in support of people. We focus on algorithms for human-robot and human-AI interaction. One of the core problems we work on is AI alignment -- getting agents (be it robots, cars, large language models, or rec systems) to do what people actually want them to by interactively learning from human feedback and modeling human decision-making. We also work on algorithms for human-AI collaboration, enabling agents to fluently assist humans in physical or virtual tasks.
1 / 6 2 / 6 3 / 6 4 / 6 5 / 6 6 / 6 Welcome to the Interactive Agents and Collaborative Technologies Laboratory at UC Berkeley! Our goal is to enable AI agents to work with, around, and in support of people. We focus on algorithms for human-robot and human-AI interaction. One of the core problems we work on is AI alignment -- getting agents (be it robots, cars, large language models, or rec systems) to do what people actually want them to by interactively learning from human feedback and modeling human decision-making. We also work on algorithms for human-AI collaboration, enabling…
saved by
related reading
- Explaining AI Alignment as an NLPer and Why I am Working on Itruiqizhong.substack.com
- Alignment remains a hard, unsolved problem — LessWronglesswrong.com
- The Future Worth Building Is Human - Thinking Machines Labthinkingmachines.ai
- Center for the Alignment of AI Alignment Centersalignmentalignment.ai
- Anca Draganpeople.eecs.berkeley.edu
- Workshop Labsworkshoplabs.ai
- Thinking Machines Labthinkingmachines.ai
- PhD_thesis_Shirley_Wu_final.pdfcs.stanford.edu
- The Artificiality of Alignmentjoinreboot.org
- Towards Bidirectional Human-AI Alignmentarxiv.org
- Patterns and problems in multiagent systemsanthropic.com
- Automated Alignment Researchers: Using large language models to scale scalable oversight \ Anthropicanthropic.com