SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2301.13019 · NeurIPS · 2023

Identifying Expert Behavior in Offline Training Datasets Improves Behavioral Cloning of Robotic Manipulation Policies

Qiang Wang, Robert Mccarthy, David Cordova Bulens, Francisco Roldan Sanchez, Kevin Mcguinness, Noel O'connor, Stephen Redmond

arXiv · PDF · Open in the Atlas

Code that ran

2 of 6 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 2 have no contract verdict; 0 failures of our instrument; 4 no recorded run), measured 2026-11-22; repository links from the Papers with Code archive attributed under CC BY-SA, no affiliation; bodies inline only under a permissive licence as GitHub detects it. Not a reproduction of the paper's results.

"Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results. This record, dated, as JSON: /api/ran/2301.13019.json · badge: /api/ran/2301.13019.svg

Each repository's role is a label: a classifier's output recorded on the link, or, where no role is recorded, what the Papers with Code archive marked. The line under it names its basis. Neither says who wrote the code.

RepositoryRoleRan
wq13552463699/real-robot-challenge-2022 canonical (classifier label)
classifier basis not recorded on the link
2 ran · 4 no recorded run
FunctionStatusWhere it lives
device_handler Ran
ran on a synthesized input
wq13552463699/real-robot-challenge-2022/utils.py
code served (permissive licence) · get_code("bc9c9a3d4d7c8a98")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Args. The image it ran in was not recorded. The code mentions cuda, and the harness aliases cuda to the CPU for every sample; whether this run needed that is not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

class Args:
    def __init__(self):
        self.use_gpu = False

ARGS = {
    'args': Args()
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Args — a Args output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Args output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
rrc_task_handler Ran
ran on a synthesized input
wq13552463699/real-robot-challenge-2022/utils.py
code served (permissive licence) · get_code("d42152a359ce3603")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Args. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import argparse

torch.manual_seed(0)

class Args:
    def __init__(self):
        self.task = None
        self.task_type = None
        self.diff = None
        self.obs_dim = None
        self.action_dim = None
        self.turns = None
        self.confs = None

# Test with 'real_lift_mix' task
args = Args()
args.task = 'real_lift_mix'

ARGS = {'args': args}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Args — a Args output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Args output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
directory_handler Not yet run
unverified (harvested, no recorded run)
wq13552463699/real-robot-challenge-2022/utils.py
code served (permissive licence) · get_code("6f510cbd5f7b96b3")
How it ran

no invocation recorded.

filter_by_net Not yet run
unverified (harvested, no recorded run)
wq13552463699/real-robot-challenge-2022/rrc_dataset_handler.py
code served (permissive licence) · get_code("47d972ac05f5d647")
How it ran

no invocation recorded.

filter_by_reward Not yet run
unverified (harvested, no recorded run)
wq13552463699/real-robot-challenge-2022/rrc_dataset_handler.py
code served (permissive licence) · get_code("358f228d782385a2")
How it ran

no invocation recorded.

load_rrc_raw_dataset Not yet run
unverified (harvested, no recorded run)
wq13552463699/real-robot-challenge-2022/rrc_dataset_handler.py
code served (permissive licence) · get_code("a1be7688817535f1")
How it ran

no invocation recorded.

Also present in other papers' code

Of 6 samples lifted from this paper's linked repositories, 6 carry a function identity and 0 do not. Of the 6 distinct functions with an identity, 0 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 0 later) and 6 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

Only here means: among the 52,977 papers whose code Syntology holds (51,887 with lifted samples, 1,090 through their assemblies only), as of 2026-09-26, no other paper's lifted samples or current assembly bricks contain an identical function body after normalisation.

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-22.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

This paper presents our solution for the Real Robot Challenge (RRC) III 1 , aiming to address dexterous robotic manipulation tasks through learning from pre-collected offline data. In this competition, participants were given two types of datasets for each task with mixed skill levels: expert and mixed. Each expert dataset is collected by a domain-specific high-skill policy, whereas the mixed dataset is collected using both expert and non-expert policies. During our early exploration of the problem, we found that the simplest offline policy learning algorithm, known as behavioural cloning (BC), exhibited remarkable learning capabilities and learned a very proficient policy with minimal human intervention when trained on expert datasets. Notably, BC outperformed even the most advanced offline reinforcement learning (RL) algorithms. However, when applied to mixed datasets, the performance of BC deteriorates; similarly, the performance of offline RL algorithms is also less than satisfactory. Upon examining the provided datasets, it was apparent that each mixed dataset contained a significant proportion of expert data, which should theoretically enable the training of a proficient BC agent. However, the expert data is not labelled in the datasets. As a result, we attempted to filter the expert data from the mixed datasets. Unfortunately, basic intuitive reward-based extraction methods were inadequate for this purpose. Therefore, we propose a classifier trained using semi-supervised learning, and designed to identify the underlying pattern of the expert behaviour within a mixed dataset. The trained classifier can then be utilized to effectively isolate the expert data. To further boost the BC performance, we take advantage of the geometric symmetry of the RRC arena to augment the training dataset through mathematical transformations. Ultimately, our submission outperformed that of all other participants, even those who employed intricate offline RL algorithms alongside complex data processing and feature engineering.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2301.13019")
get_code_for_paper("2301.13019")
have("2301.13019")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/2301.13019.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/2301.13019.svg)](https://syntology.ai/paper/2301.13019)

Connect an agent — have() is free.