SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2210.07442 · 2022

Frame Mining: a Free Lunch for Learning Robotic Manipulation from 3D Point Clouds

arXiv · PDF · Open in the Atlas

Code that ran

5 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 5 have no contract verdict; 0 failures of our instrument; 1 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/2210.07442.json · badge: /api/ran/2210.07442.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
xuanlinli17/corl_22_frame_mining canonical (classifier label)
classifier basis not recorded on the link
5 ran · 1 no recorded run
FunctionStatusWhere it lives
combine_obs_with_action Ran
ran on a synthesized input
xuanlinli17/corl_22_frame_mining/pyrl/pyrl/networks/utils.py
code served (permissive licence) · get_code("634131bf78bb67ee")
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 dict. 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

torch.manual_seed(0)

# Test case 1: obs is a tensor, action is a tensor
obs_tensor = torch.randn(2, 4)
action_tensor = torch.randn(2, 3)

# Test case 2: obs is a dict with 'state', action is a tensor
obs_dict = {
    'state': torch.randn(2, 4),
    'other_key': 'some_value'
}
action_for_dict = torch.randn(2, 3)

# Test case 3: obs is a tensor, action is None
obs_tensor_no_action = torch.randn(2, 4)

# Create ARGS for the primary use case: dict obs with action
ARGS = {
    'obs': obs_dict,
    'action': action_for_dict
}
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 dict — a dict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a dict 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
concat_vec_in_dict Ran
ran on a synthesized input
xuanlinli17/corl_22_frame_mining/mani_skill/mani_skill/agent/agent.py
code served (permissive licence) · get_code("4de128b63b27977a")
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 ndarray, ndim 1, dtype family float, shape (6,) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape 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 numpy as np

# Create a sample dictionary with various types of values
d = {
    'array_key': np.array([1.0, 2.0, 3.0]),
    'scalar_key': 5,
    'another_array': np.array([4.0, 5.0]),
    'unused_key': np.array([99.0]),
}

# Create a list of keys to concatenate
key_list = ['array_key', 'scalar_key', 'another_array']

ARGS = {
    'd': d,
    'key_list': key_list,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 1, dtype family float, shape (6,) — shape (6,) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
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
get_actor_by_name Ran
ran on a synthesized input
xuanlinli17/corl_22_frame_mining/mani_skill/mani_skill/agent/agent.py
code served (permissive licence) · get_code("101a97ceba07b65b")
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 Actor. 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.

class Actor:
    def __init__(self, name):
        self._name = name
    
    def get_name(self):
        return self._name

actor1 = Actor("alice")
actor2 = Actor("bob")
actor3 = Actor("charlie")

actors = [actor1, actor2, actor3]
names = "bob"

ARGS = {
    "actors": actors,
    "names": names
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
none outside the standard library — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Actor — a Actor output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Actor 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
get_kwargs_from_shape Ran
ran on a synthesized input
xuanlinli17/corl_22_frame_mining/pyrl/pyrl/networks/utils.py
code served (permissive licence) · get_code("3f77c48560c629db")
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 dict. 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 numpy as np
from copy import deepcopy

# Test case 1: dict obs_shape with pointcloud and state
obs_shape_dict = {
    'state': (8,),
    'hand_pose': (7, 4, 2),  # last dim -2 is 4
    'pointcloud': {
        'xyz': (1024, 3),
        'rgb': (1024, 3),
        'seg': (1024, 5),
        'target_object_point': 1
    }
}

action_shape = (4,)

ARGS = {
    'obs_shape': obs_shape_dict,
    'action_shape': action_shape
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type dict — a dict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a dict 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
parse_level_range Ran
ran on a synthesized input
xuanlinli17/corl_22_frame_mining/mani_skill/evaluate_policy.py
code served (permissive licence) · get_code("806a2ec8e223f968")
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 tuple, ndim 1, dtype family int; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim 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 torch.nn as nn
import numpy as np

# Test case 1: Valid level range string
s_valid = "100-200"

# Test case 2: None input
s_none = None

# Test case 3: Invalid format (no dash) - would raise exception
s_invalid = "100200"

# For the function, we need a single parameter 's' which is a string or None
# The function splits on '-' and converts to integers
# Valid cases: "100-200" or None
# Invalid case: string without '-'

ARGS = {
    "s": "100-200"
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple, ndim 1, dtype family int — a tuple output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
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
build Not yet run
unverified (harvested, no recorded run)
xuanlinli17/corl_22_frame_mining/pyrl/pyrl/networks/builder.py
code served (permissive licence) · get_code("053001fc2285043b")
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, 1 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 1 later) and 5 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.

FunctionAlso present inEarliest-datedOthers named
parse_level_range 1 other paper
1 later
2306.06799 (2023-06)
On the Efficacy of 3D Point Cloud Reinforcement Learning

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

We study how choices of input point cloud coordinate frames impact learning of manipulation skills from 3D point clouds. There exist a variety of coordinate frame choices to normalize captured robot-object-interaction point clouds. We find that different frames have a profound effect on agent learning performance, and the trend is similar across 3D backbone networks. In particular, the end-effector frame and the target-part frame achieve higher training efficiency than the commonly used world frame and robot-base frame in many tasks, intuitively because they provide helpful alignments among point clouds across time steps and thus can simplify visual module learning. Moreover, the well-performing frames vary across tasks, and some tasks may benefit from multiple frame candidates. We thus propose FrameMiners to adaptively select candidate frames and fuse their merits in a task-agnostic manner. Experimentally, FrameMiners achieves on-par or significantly higher performance than the best single-frame version on five fully physical manipulation tasks adapted from ManiSkill and OCRTOC. Without changing existing camera placements or adding extra cameras, point cloud frame mining can serve as a free lunch to improve 3D manipulation learning.

For agents

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

get_harvested_code_for_paper("2210.07442")
get_code_for_paper("2210.07442")
have("2210.07442")

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

curl https://syntology.ai/api/ran/2210.07442.json

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

[![Syntology run record](https://syntology.ai/api/ran/2210.07442.svg)](https://syntology.ai/paper/2210.07442)

Connect an agent — have() is free.