EmbodiedSWE: Coding Agents for Long Horizon Dexterous Robotics

Zeyu Shen; Haoxiang You; Yilang Liu; Zhicheng Zheng; Lihan Zha; Kashu Yamazaki; Mingtong Zhang; Suning Huang

cs.ROSep 23, 2026

EmbodiedSWE: Coding Agents for Long Horizon Dexterous Robotics

Authors: Zeyu Shen, Haoxiang You, Yilang Liu, Zhicheng Zheng, Lihan Zha, Kashu Yamazaki, Mingtong Zhang, Suning Huang,

Organizations: ByteDance Seed · Princeton University · Yale University · Carnegie Mellon University ·

Abstract

We study coding agents for long-horizon, dexterous robotics and ask whether their solutions can provide scalable supervision for learning general robot policies. To test this, we develop EMBODIEDSWE-BENCH, a simulation benchmark for coding agents spanning contact-rich manipulation, deformable objects, and long-horizon tasks requiring up to half an hour of continuous interaction. We find that frontier coding agents can solve complex long-horizon tasks and transfer prior solutions across both tasks and embodiments. We also design supporting tools that help agents more effectively solve these tasks. However, the resulting solutions require substantial iterative interaction and are typically specialized to individual task instances. We therefore introduce EMBODIEDSWE-GEN, which expands a single solution from coding agent into large diverse trajectories for training a VLA. VLA performance improves with more generated demonstrations, and agent-aided diversification improves generalization to held-out task variations. We also show that a VLA finetuned solely on coding-agent-generated simulation demonstrations completes a long-horizon task on real robot. Together, our framework uses coding agents to solve complex robotics tasks and turn verified solutions into scalable supervision for robot policies.

Figures & tables

Figure 1: Overview of our framework: Top: EmbodiedSWE-Bench contains diverse long-horizon robotics tasks. Bottom: coding agents solve these tasks and turn verified solutions into scalable supervision for general policies.

Figure 2: Representative successful trajectories from coding-agent solutions on EmbodiedSWE-Bench , spanning long-horizon manipulation across tasks and embodiments.

Figure 3: Base results on EmbodiedSWE-Bench . Left, middle: mean best-so-far score over 28 tasks against wall-clock and tokens; markers with vertical bars at hourly intervals / fixed token counts show mean

\pm

SEM across tasks (one run per task). Right: mean final score with SEM across tasks; percentages give the fraction of tasks solved.

Figure 4: Score against cost.

Figure 5: Programmatic policies versus RL. Despite stronger supervision and extensive tuning, RL substantially underperforms coding-agent policies.

Figure 6: Evaluations on cross-task and cross-embodiment transfer (left two) and with additional tools (right two). Markers with vertical bars at hourly intervals show mean

\pm

SEM across runs (2 models

\times

3 seeds per task for the generalization runs; one run per task for the tool runs).

Figure 7: Hierarchical diversification pipeline of EmbodiedSWE-Gen . Starting from one verified coding-agent solution, we diversify the scene , strategy , phase , dynamics , and visual to expand it into a broad set of training trajectories. Higher-level variations modify task structure, solution behavior, and intermediate-state coverage, while lower-level variations introduce physical, action, and visual perturbations to improve robustness and diversify execution.

Figure 8: One verified solution can be expanded into many diverse trajectories through hierarchical diversification, with only modest additional agent cost per variation and rapidly decreasing amortized cost at scale.

Figure 9: VLA finetuning with generated in-distribution demonstrations. Mean task scores and success rates at five training-set sizes. Overall performance improves as more generated demonstrations are provided.

Task	Script-only	Agent-aided (ours)
pen_holder	$0.083\pm 0.040$	$0.183\pm 0.021$
tool_packing	$0.262\pm 0.028$	$0.331\pm 0.026$
slice	0	$0.220\pm 0.017$
bulb	$0.026\pm 0.013$	$0.236\pm 0.081$
nut_thread	$0.025\pm 0.006$	$0.300\pm 0.027$
clear_organic_objects	$0.002\pm 0.001$	$0.129\pm 0.023$

Table 1: Generalization to held-out variations. Both methods use the same number of training demonstrations.

Figure 10: Successful real-robot lamp disassembly rollout from a policy finetuned only on coding-agent-generated simulation data

Policy	Grasp shade	Place and release shade	Grasp bulb	Unscrew and Extract bulb
Sim-trained	100%	80%	30%	20%
Pretrained $\pi_{0.5}$ -DROID	0%	0%	0%	0%

Table 2: Real-robot stage success rates over 10 trials. The finetuned policy achieves nontrivial success on the real robot, including completion of the challenging bulb-unscrewing stage, despite being trained exclusively on simulated demonstrations.

Appendix figures & tables25 assets

Supplementary material from the paper’s appendix.

Appendix

Suite	Task	Supported embodiments	Difficulty
Assembly	allen_bolt . Drive an Allen bolt into a threaded plate using an L-shaped key.	Franka; UFACTORY xArm7; Kinova Gen3	Medium
Assembly	bulb . Pick up a light bulb and screw it into its socket until fully seated.	Franka; UFACTORY xArm7; Kinova Gen3	Medium
Assembly	ikea_table . Attach four threaded legs to the corner studs of a tabletop.	Franka (bimanual)	Hard
Assembly	SO_101 . Assemble a robot arm.	Franka (bimanual)	Hard
Assembly	nut_thread . Pick up an M16 nut and thread it onto a fixed bolt.	Franka; UFACTORY xArm7; Kinova Gen3	Easy
Assembly	pc_gpu . Align a graphics card with a PCIe slot and press it into place.	Franka; UFACTORY xArm7; Kinova Gen3	Easy

Appendix

Table 3: EmbodiedSWE-Bench task catalog.

Type	Embodiment	Arm DOFs	End-effector	Notes
	Franka	7	two-finger parallel gripper	fixed base; tabletop manipulation
Single arm	UFACTORY xArm7	7	linkage gripper or Panda hand	$\sim$ 0.7 m reach
	Kinova Gen3	7	Panda hand at the flange	$\sim$ 0.9 m reach
Bimanual	Bimanual Franka	14	two parallel grippers	two Frankas with per-scene base placement; actions concatenated
Humanoid	Unitree G1	17	two three-finger hands (14 DOFs)	29-DOF humanoid; pelvis fixed in tabletop tasks; in locomanipulation tasks, the 12 leg DOFs are driven by a pretrained locomotion policy taking a base velocity command

Appendix

Table 4: Supported embodiments.

Controller	Commands	Action (per arm)	Used in
Operational-space control	torques (actuators at zero intrinsic stiffness)	six end-effector pose deltas	arm evaluations (default)
Differential IK	joint position targets tracked by the arm PD	six end-effector pose deltas	arm evaluations
Pink IK	joint position targets tracked by the arm PD	absolute end-effector pose	humanoid evaluations
Task-space impedance	torques (actuators at zero intrinsic stiffness)	six end-effector pose deltas	—
Direct joint targets	joint position targets	one target per joint	—

Appendix

Table 5: Provided controllers.

Figure 11: Best-so-far score versus wall-clock time, by difficulty tier.

Figure 12: Best-so-far score versus generated tokens, by difficulty tier.

Figure 13: Best-so-far score versus wall-clock time, by suite.

Figure 14: Best-so-far score versus generated tokens, by suite.

Figure 15: Best-so-far score versus wall-clock time, by embodiment.

Figure 16: Best-so-far score versus generated tokens, by embodiment.

Task	Stage terms (weight)
bulb	reach (0.1): pads at the bulb’s glass belly; lift (0.1): rise above the start height, full at 8 cm; transport (0.2): kernel on the bulb’s lateral offset from the nearest socket axis; upright (0.1): cosine between the bulb and socket axes; approach (0.2): when aligned, kernel on the height above the bore mouth; thread (0.3): when engaged, depth from the bore mouth to the seat.
nut_thread	reach (0.1): pads at the nut; lift (0.1): rise above the table, full at 6 cm; transport (0.2): lateral offset from the nearest bolt axis; upright (0.1): cosine between the nut and bolt axes; approach (0.2): when aligned, height above the thread start; thread (0.3): when engaged, depth from the thread start to the seat.
slice	reach (0.1): pads at the knife handle; taken (0.1): knife rise off its rest, full at 6 cm; carry (0.15): blade-edge centre over the food; align (0.15): blade normal parallel to the cut axis; press (0.2): the edge sample nearest the next uncut plane lies in that plane and is pressed toward the board; cut (0.3): fraction of planes released.
pen_holder	reach (0.1): pads at the nearest not-yet-inserted pen; lift (0.1): that pen’s rise off the table, full at 10 cm; carry (0.15): the pen’s bottom end over the holder axis; tip_up (0.1): pen axis along the holder axis, tip up; insert (0.25): when over the axis, pen bottom lowered below the rim toward the target depth; all_in (0.3): fraction of pens counted as inserted.
tool_packing	reach (0.1): pads at the nearest unstowed tool; lift (0.1): its rise off the surface, full at 8 cm; carry (0.2): tool over its assigned drawer; stowed (0.3): fraction of tools inside their drawers; drawers (0.2): mean drawer shut-ness; doors (0.1): mean door shut-ness.

Appendix

Table 15: Agent-designed dense rewards. Final stage terms and weights proposed by the coding agent for each task. Every term lies in

[0,1]

; the potential is the weighted sum normalized by the total weight. Reach terms use a two-scale kernel

\tfrac{1}{2}[(1-\tanh(d/0.30))+(1-\tanh(d/0.05))]

on the distance

d

between the finger pads and the grasp point; ramps are clipped to

[0,1]

Task	Stage terms	Training weights
bulb	reach (grasp pose at the neck waist), grasp, lift, transport, upright, approach, thread	reach 0.2, grasp 0.2, lift 0.6
nut_thread	reach (grasp pose on the upper half of the nut), grasp, lift, transport, upright, approach, thread	reach 0.2, grasp 0.2, lift 0.6; resumed with reach 0.1, grasp 0.1, lift 0.2, transport 0.3, upright 0.1, approach 0.2
slice	reach (grasp pose at the handle centre, fingers across the plate), grasp, taken, carry, align, press, cut	reach 0.2, grasp 0.2, taken 0.6
pen_holder	reach (nearest uninserted pen), grasp, lift, carry, tip_up, insert, all_in	reach 0.2, grasp 0.2, lift 0.6
tool_packing	drawer ladder: reach_h (handle pinch pose), grasp_h, open (pull-out ramp plus bonus); item ladder, gated on the drawer being open: reach, grasp, lift, orient, carry, stowed	reach_h 0.05, grasp_h 0.05, open 0.15, reach 0.1, grasp 0.15, lift 0.2, orient 0.05, carry 0.1, stowed 0.15

Appendix

Table 16: Expert-tuned rewards. Stage terms defined by the expert for each task and the weights used in the final training runs. Because no policy learned the pick under a flat weighting (on pen_holder , grasp peaked at

0.04

and lift at

0.0003

over 300 iterations with seven terms at roughly

0.1

each), the weight was concentrated on the pick terms for the tasks whose first rubric stage is a lift; the remaining terms are defined and logged but weighted zero. tool_packing uses a per-item ladder in which the pick-and-place terms are gated on the assigned drawer being open.

Task	Additions
bulb	Grasp point moved from the glass belly to the neck waist (2.5 cm above the cap, 2 cm wide); pad offset re-measured; warm-start hover target 10 cm above the neck.
nut_thread	Pinch the upper half of the 13 mm nut (24 mm across flats); pad offset re-measured to the fingertips; warm-start hover target 10 cm above the nut; final run initialized from the bulb policy.
slice	Joint-delta actions with scale $0.15$ rad per step (the arm PD caps joint speed at scale $\cdot k_{p}/k_{d}$ , and $0.05$ was too slow to reach the knife); lift target raised to 8 cm because the knife settles 2.1 cm into its notches after reset while the grader keeps the unsettled height; warm-start hover target 10 cm above the handle, with a bridging variant that stops 60% of the warm-started environments partway along the home-to-hover path; episode terminates if the knife falls off the table.
pen_holder	Finger targets limited to 0–20 mm with a first-order filter (coefficient 0.9), since a single noisy finger action dropped an in-hand pen; measured 7.5 mm hold width (the collider is far thinner than the visual barrel); warm-start hover target above the focused pen, with optional in-hand and re-yawed pen starts.
tool_packing	A drawer-pull ladder (reach handle, grasp handle, pull-out ramp with a bonus once 7 cm of tray is exposed) placed before the per-item pick-and-drop ladder; the stapler as the single active item; warm-start hover target in front of the drawer handle, or above the item once its drawer is open; a curriculum that resets half of the environments with the assigned drawer already open.

Appendix

Table 17: Per-task environment changes in the expert-tuned condition. Measured quantities were obtained by probing the simulator. Curricula apply to training only; grading always starts from the home pose.

Task	Horizon	Steps	Ent.	Finger $k_{p}$	Warm start	Jitter (cm)
bulb	14	32	0.006	8,000	50%, 90	3
nut_thread	14	32	0.006	8,000	50%, 90	3
slice	20	64	0.006	2,000	60%, 100	2
pen_holder	14	32	0.006	2,000	50%, 90	3
tool_packing	24	32	0.006	8,000	50%, 90	3

Appendix

Table 18: Fine-tuned training hyperparameters in the expert-tuned condition. PPO settings not listed keep the values in the text. The warm-start curriculum pre-rolls the given fraction of environments to the hover target for the given number of control steps, with lateral jitter, before each episode.

Target	Similar source	Shared skill	Dissimilar source
bulb	nut_thread	screwing along a thread	tool_packing
allen_bolt	nut_thread	threading/ratcheting a screw	pen_holder
pc_motherboard	allen_bolt	hex-key bolt ratcheting	spatula
pc_ram	pc_gpu	seating a card in a slot	bulb
tool_packing	pen_holder	stowing objects into recesses	nut_thread
spatula	allen_bolt	wielding a grasped tool	pc_gpu

Appendix

Table 19: Cross-task transfer pairs. Similar sources share the target’s dominant contact skill; dissimilar sources have a different dominant skill.

Experiment	Setting	Score	Success	Tokens per run (M)
Experiment	Setting	Score	Success	Unc. input	Cached input	Output
Cross-task	similar-task hint	$0.60\pm 0.05$	31%	1.9	54	0.13
	dissimilar-task hint	$0.51\pm 0.05$	17%	1.9	86	0.15
	no hint	$0.52\pm 0.05$	11%	2.3	84	0.15
Cross-embodiment	Kinova Gen3 + hint	$0.74\pm 0.05$	56%	0.8	56	0.10
	xArm7 + hint	$0.66\pm 0.06$	36%	1.3	55	0.10

Appendix

Table 20: Transfer learning results. Final rubric score (mean

\pm

SEM over 36 runs), success rate (fraction of runs ending at rubric score

1.0

), and mean token use per run in millions. Both no-hint rows use the same from-scratch Franka runs on the six target tasks.

Figure 17: Cross-task generalization by target task. Rubric score versus agent wall-clock time (mean

\pm

SEM over 2 models

\times

3 seeds) for the similar-task hint, dissimilar-task hint, and no-hint settings.

Figure 18: Cross-embodiment generalization by task. Rubric score versus agent wall-clock time (mean

\pm

SEM over 2 models

\times

3 seeds) for the two target embodiments given the Franka reference solution as a hint. From-scratch Franka runs are shown as a reference for task difficulty.

Success rate (%)
	Number of demonstrations
Task	10	50	100	200	400
coffee	$3.10\pm 3.10$	$6.20\pm 4.30$	$25.00\pm 7.70$	$31.00\pm 8.20$	$34.00\pm 8.40$
pc_gpu	$0.00\pm 0.00$	$41.00\pm 12.00$	$34.00\pm 21.00$	$69.00\pm 15.00$	$81.00\pm 11.00$
clear_organic_objects	$16.00\pm 24.00$	$34.00\pm 28.00$	$41.00\pm 6.00$	$41.00\pm 12.00$	$50.00\pm 18.00$
pen_holder	$3.00\pm 3.00$	$49.00\pm 9.00$	$34.00\pm 8.00$	$50.00\pm 9.00$	$59.00\pm 9.00$

Appendix

Table 21: VLA finetuning results. Success rates (%) and task scores at five training-set sizes. Revised success criteria for selected tasks are listed in Table 22 .

Task	Revised success criterion
Coffee preparation	Press START with the bay cover closed, the capsule seated, and the mug positioned under the spout.
Clearing organic objects	Complete the placement of one organic object.
Pen holding	Insert one pen into the holder.
Banana slicing	Cut the banana into at least three pieces.

Appendix

Table 22: Task-specific evaluation criteria. For some tasks, we relax the original success conditions as follows.

Panel of Figure 19 — Figure 19: Illustration of the real setup versus the constructed sim setup.

Figure 20: RL training curves.

Figure 21: Failure mode 1a: Opus 5 hallucinates progress on Syringe task.

Figure 22: Failure mode 1b: Opus 5 fails to notice pathological states on the pen_holder task.

Figure 23: Failure mode 2: Opus 4.8 over-commits to a failing plan on the Syringe task.

Figure 24: Failure mode 3: GPT-5.6-Sol hacks the scorer on the Syringe task.

Coding Agents

Coding agents use language models, tools, and feedback to modify software and complete engineering tasks.

See all 141 papers

RLE-Bench: A Qualifying Exam for Coding Agents as Robot Learning Engineers
Sep 28, 2026Haitong Ma, Chenxiao Gao, Rushi Qiang +2AI Agent Benchmarks Robot Learning
Harvard University · Georgia Institute of Technology
ENPIRE: Agentic Robot Policy Self-Improvement in the Real World
Jun 18, 2026Wenli Xiao, Jia Xie, Tonghe Zhang +14Robot Policy Learning Robot Skill Learning
1NVIDIA · 2CMU · 3UC Berkeley
EmbodiedEvo: Self-Evolving Coding Agent For Physical Intelligence
Oct 13, 2026Wenyi Li, Yizhe Chi, Youjie Zheng +8Self-Improving Agents Robot Skill Learning
Navers Lab, Einsia.AI, Tsinghua University

Sep 28, 2026J/K move · Enter open · S save

RLE-Bench: A Qualifying Exam for Coding Agents as Robot Learning Engineers

Haitong Ma, Chenxiao Gao, Rushi Qiang, Bo Dai+1

Harvard University · Georgia Institute of Technology

AI Agent Benchmarks Robot Learning

Open paper AskPDF arXiv

Coding agents are beginning to move beyond purely digital tasks to tackle physical-world challenges, particularly in robotics. Existing robotics benchmarks, however, primarily focus on the performance of individual artifacts, such as policies or controllers, offering limited coverage of coding agents' broader engineering capabilities. Real-world robotics extends beyond control: agents must build, integrate, diagnose, and improve heterogeneous artifacts under resource constraints and reason from multimodal feedback. To evaluate these broader capabilities, we introduce RLE-Bench, a benchmark of robot-learning tasks spanning four representative robotics development workflows: interactive control, policy learning, perception and estimation, and mechanical design. We use diverse task-specific metrics to evaluate the artifacts submitted by the coding agents, from the success rate the agents achieved to the policy agents trained, the harness agent built, and the mechanical structures the agent designed. We aggregate these metrics into an overall RLE Index and report workflow-specific capability profiles, enabling systematic comparison of coding agents' capabilities across multiple capability dimensions. Beyond performance ranks, we also conduct in-depth case studies examining agent behavior on representative tasks, highlighting both current capabilities and limitations, and pointing to the opportunities robotics tasks have to offer for future agent training.

Abstract

Coding Agents

Explore similar work

RLE-Bench: A Qualifying Exam for Coding Agents as Robot Learning Engineers

ENPIRE: Agentic Robot Policy Self-Improvement in the Real World

EmbodiedEvo: Self-Evolving Coding Agent For Physical Intelligence