Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Concept Videos

Stereotype Content Model02:16

Stereotype Content Model

13.9K
The Stereotype Content Model (SCM) was first proposed by Susan Fiske and her colleagues (Fiske, Cuddy, Glick & Xu, 2002; see also Fiske, 2012 and Fiske, 2017). The SCM specifies that when someone encounters a new group, they will stereotype them based on two metrics: warmth—or that group’s perceived intent, and how likely they are to provide help or inflict harm—and competence—or their ability to carry out that objective. Depending on the warmth-competence...
13.9K

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Lossy encoding of distributions in judgment under uncertainty.

Cognitive psychology·2025
Same author

Automating the practice of science: Opportunities, challenges, and implications.

Proceedings of the National Academy of Sciences of the United States of America·2025
Same author

Beyond discrete-choice options.

Trends in cognitive sciences·2024
Same author

Designing optimal behavioral experiments using machine learning.

eLife·2024
Same author

Naïve information aggregation in human social learning.

Cognition·2023
Same author

Local Search and the Evolution of World Models.

Topics in cognitive science·2023
Same journal

Algorithm-hardware co-design of neuromorphic networks with dual memory pathways.

Nature machine intelligence·2026
Same journal

Plagiarism in the Age of Generative Artificial Intelligence: The advent of generative artificial intelligence (GenAI) tools is challenging the scientific community's understanding of the meaning and significance of plagiarism. A new definition of research misconduct is needed that specifically addresses the use of GenAI writing tools.

Nature machine intelligence·2026
Same journal

Platonic representation of foundation machine learning interatomic potentials.

Nature machine intelligence·2026
Same journal

Immunotherapy drug target identification using machine learning and patient-derived tumour explant validation.

Nature machine intelligence·2026
Same journal

A generative artificial intelligence approach for peptide antibiotic optimization.

Nature machine intelligence·2026
Same journal

LLMs displaying less cognitive bias are not necessarily better decision makers.

Nature machine intelligence·2026
See all related articles

Related Experiment Video

Updated: May 22, 2025

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness
03:14

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness

Published on: December 6, 2024

474

Embodied large language models enable robots to complete complex tasks in unpredictable environments.

Ruaridh Mon-Williams1,2,3, Gen Li1, Ran Long1

  • 1University of Edinburgh, Edinburgh, UK.

Nature Machine Intelligence
|May 20, 2025
PubMed
Summary
This summary is machine-generated.

Robots can now complete complex, long-horizon tasks in unpredictable environments using the new embodied large-language-model-enabled robot (ELLMER) framework. This AI-powered system integrates artificial intelligence with robotic sensorimotor skills for adaptive task completion.

Keywords:
Computer scienceEngineering

More Related Videos

Haptic/Graphic Rehabilitation: Integrating a Robot into a Virtual Environment Library and Applying it to Stroke Therapy
13:44

Haptic/Graphic Rehabilitation: Integrating a Robot into a Virtual Environment Library and Applying it to Stroke Therapy

Published on: August 8, 2011

13.8K
The Modular Design and Production of an Intelligent Robot Based on a Closed-Loop Control Strategy
11:53

The Modular Design and Production of an Intelligent Robot Based on a Closed-Loop Control Strategy

Published on: October 14, 2017

11.5K

Related Experiment Videos

Last Updated: May 22, 2025

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness
03:14

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness

Published on: December 6, 2024

474
Haptic/Graphic Rehabilitation: Integrating a Robot into a Virtual Environment Library and Applying it to Stroke Therapy
13:44

Haptic/Graphic Rehabilitation: Integrating a Robot into a Virtual Environment Library and Applying it to Stroke Therapy

Published on: August 8, 2011

13.8K
The Modular Design and Production of an Intelligent Robot Based on a Closed-Loop Control Strategy
11:53

The Modular Design and Production of an Intelligent Robot Based on a Closed-Loop Control Strategy

Published on: October 14, 2017

11.5K

Area of Science:

  • Robotics and Artificial Intelligence
  • Machine Learning
  • Embodied AI

Background:

  • Complex tasks in unpredictable environments challenge current robotic systems, necessitating advancements in machine intelligence.
  • Sensorimotor abilities are crucial for human intelligence, suggesting their importance for biologically inspired machine intelligence.
  • Integrating artificial intelligence (AI) with robotic sensorimotor capabilities offers a promising direction for developing advanced robots.

Purpose of the Study:

  • To introduce an embodied large-language-model-enabled robot (ELLMER) framework designed for complex, long-horizon tasks in unpredictable settings.
  • To enable robots to adapt to changing conditions by incorporating force and visual feedback into action plans.
  • To demonstrate the capability of AI-driven robots to perform intricate tasks requiring sequential sub-tasks and diverse feedback mechanisms.

Main Methods:

  • Utilized GPT-4 and a retrieval-augmented generation infrastructure within the ELLMER framework.
  • Extracted contextually relevant examples from a knowledge base to generate action plans.
  • Incorporated force and visual feedback for adaptive control and task execution.

Main Results:

  • The ELLMER framework successfully enabled a robot to complete complex tasks, including coffee making and plate decoration.
  • Demonstrated the robot's ability to execute a sequence of sub-tasks, such as opening drawers and pouring, with distinct feedback.
  • Showcased the framework's effectiveness in adapting to changing environmental conditions during task execution.

Conclusions:

  • The ELLMER framework represents significant progress towards creating scalable, efficient, and intelligent robots.
  • This approach facilitates robots capable of performing complex tasks in uncertain and dynamic environments.
  • The integration of large language models and sensorimotor feedback is a key advancement in embodied AI.