This paper introduces a novel Balance Feedback Interface (BFI) that addresses the problem of bilateral feedback for teleoperation of humanoid robots. With this new device we expect to enhance robot's high force manipulation performance to a level comparable to humans by dynamically synchronizing master and slave. Through this Human-Machine Interface (HMI) we aim to achieve two goals: (i) have the human pilot learn from the robot's dynamic behavior; and (ii) have the humanoid robotic platform learn from human's motor skills. The eventual goal is to fuse the teleoperator's commands and an autonomous controller for optimal performance. Initial results evaluate the stability of the robot while being teleoperated by a human pilot for a simple upright balancing task. During the experiment the user has no visual information about the robot's state and is expected to compensate for unpredicted instabilities. The bilateral feedback system shows robustness to inertial variations and to external disturbances with the proposed human-in-the-loop control strategy offering valuable insight for future work.