Skip to content
Agientry
← Back to Videos

KeyFlow Uses Coding Agents to Adapt Retrieval Policies

Loading video
Loading video

MIT LISLab introduces KeyFlow, where a coding agent improves a retrieval-based policy through simulated play so robots adapt from a few demonstrations without retraining.

Robot LearningRetrieval PolicyCoding AgentSimulation
Category: research
Author: @kyiiiifan
Date: 2026-11-05T00:00:00
Duration: 35.0s
Source: https://x.com/kyiiiifan/status/2118405829733085190
Reference: https://yifank.github.io/keyflow/

Related Videos

GPT-6 Runs Real2Sim End to End from Robot Demonstrations
0:17

GPT-6 Runs Real2Sim End to End from Robot Demonstrations

@Lingxiao234 · 1.6k views · 2026-09-07
Real-to-SimGPT-6
Perceptron Cuts Robot Video Annotation Cost to One-Eighth
0:21

Perceptron Cuts Robot Video Annotation Cost to One-Eighth

@XRoboHub · 38 views · 2026-11-04
Robot LearningEgocentric Video
Genie 4: A Steerable World Model for Robot Training
0:15

Genie 4: A Steerable World Model for Robot Training

@TheHumanoidHub · 157 views · 2026-10-27
Genie 4World Models
MessyMem: Persistent Spatial Memory for Robots
0:48

MessyMem: Persistent Spatial Memory for Robots

@anuvabanwasi · 953 views · 2026-09-16
Persistent MemoryMobile Manipulation
Lumos NexCore: End-to-End Robot Skill Platform
0:56

Lumos NexCore: End-to-End Robot Skill Platform

@LumosRobotics · 736 views · 2026-09-15
Robot SkillsRobot Learning
Odyssey-3: A Foundation World Model for Physical Agents
1:37

Odyssey-3: A Foundation World Model for Physical Agents

@odysseyml · 1.3k views · 2026-09-15
World ModelsFoundation Models

As an Amazon Associate, we earn from qualifying purchases.