Training-free Behavior Cloning

Maximilian Adang, Timothy Chen, Lars Osterberg, Aiden Swann, Mac Schwager · N/A · 2026

Summary

Abstract Summary

Neural behavior cloning compresses demonstrations into large models, making individual actions difficult to trace and policy updates costly. Retrieval policies retain access to demonstrations but struggle with mismatch between recorded and live behavior. We introduce Behavior Predictive Control (BPC), which synthesizes policies without end-to-end policy training by combining an action-aware retrieval metric, a Hankel-based action-continuation prior, and a closed-form one-step residual correction. Inspired by behavioral systems theory, BPC predicts future actions by blending stored observation-action data that best reconstructs the recent runtime observation--action history. Across simulated benchmarks and real-robot deployments, BPC is competitive with learned policies such as $π_{0.5}$ (surpassing it in some cases), while reducing policy fitting from hours to seconds on consumer GPUs and supporting closed-loop control upwards of 75 Hz on a Jetson Orin Nano. The retrieved demonstration windows and their coefficients also provide an intrinsic estimate of task progress. Retaining demonstrations within the deployed policy makes its predictions traceable to supporting trajectories and enables behavior revision through the demonstration bank.

Key Points

Neural behavior cloning compresses demonstrations into large models, making individual actions difficult to trace and policy updates costly.
Retrieval policies retain access to demonstrations but struggle with mismatch between recorded and live behavior.
We introduce Behavior Predictive Control (BPC), which synthesizes policies without end-to-end policy training by combining an action-aware retrieval metric, a Hankel-based action-continuation prior, and a closed-form one-step residual correction.
Inspired by behavioral systems theory, BPC predicts future actions by blending stored observation-action data that best reconstructs the recent runtime observation--action history.
Across simulated benchmarks and real-robot deployments, BPC is competitive with learned policies such as $π_{0.5}$ (surpassing it in some cases), while reducing policy fitting from hours to seconds on consumer GPUs and supporting closed-loop control upwards of 75 Hz on a Jetson Orin Nano.

Regularized Local Action Continuation

We adapt the trajectory-completion problem of De Persis and Tesi [ 15 , Sec. II-A, Problem 1] to behavior cloning: demonstration data constitutes a library of behaviors. Rather than predicting future observations based on a history of observations and executed actions, we seek to predict the expert action that should follow the observed history.

Time indices use parenthesized superscripts; library-column and demonstration indices use subscripts.

Regularized Local Action Continuation — Training-free Behavior Cloning — Fig. 2 : Online BPC pipeline. (1) A learned metric retrieves $K$ demonstration windows from the live observation–action history $z$. (2) Regularized reconstruction fits coefficients $g_{i}$ to their past histories and applies them to future action blocks to form the action prior. (3) Random Fourier features of retrieved observations, actions, and observation differences from the query are uniformly pooled. A linear head predicts the correction $\delta u^{(t)}$. The action prior is added to the correction to form the outgoing control action.

Extension to Nonlinear Behaviors

This perspective motivates two complementary components: future-informed retrieval to select the support for local continuation, and a shared-support nonlinear residual to correct errors that remain.

Because the action sensitivity $J_{z}$ is unknown, we use demonstrated continuations to fit a retrieval metric that favors histories predictive of valid future actions, while requiring only the observed history at deployment.

Extension to Nonlinear Behaviors — Training-free Behavior Cloning — TABLE II : Tabletop manipulation hardware: qualitative task sequences and successes over number of trials.

Results

We test BPC on several hardware manipulation tasks using the UFactory xArm6. Full task descriptions are as follows: Coffee : Pick a bag of beans from a 30 cm by 30 cm area and place it on the drawer. Drawer : Open the drawer using the small handle, pick the sponge, place the sponge in the drawer, and then close the drawer.

Switch : Pick the bag of beans and place it on the opposite plate (either the left or right side).

Results — Training-free Behavior Cloning — TABLE III : Success rates (%) on simulation tabletop manipulation tasks over 300 rollouts per task.

Sources and Demonstrations

Figures are reproduced from Adang et al. See the full paper for experimental details and the project page, when the paper links one, for demonstrations and videos.