SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2004.13242 · 2020

Efficient Black-Box Planning Using Macro-Actions with Focused Effects

arXiv · PDF · Open in the Atlas

Code that ran

8 of 13 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 8 have no contract verdict; 0 failures of our instrument; 5 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/2004.13242.json · badge: /api/ran/2004.13242.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
camall3n/focused-macros canonical (classifier label)
classifier basis: census
5 ran · 4 no recorded run
camall3n/skills-for-planning canonical (classifier label)
classifier basis: census
3 ran · 1 no recorded run
FunctionStatusWhere it lives
as_range Ran
ran on a synthesized input
camall3n/focused-macros/experiments/show_missing.py
code served (permissive licence) · get_code("c8a878ab793e2340")
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 str. 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 numpy as np

def as_range(iterable):
    list_ = list(iterable)
    if len(list_) > 1:
        return '{0}-{1}'.format(list_[0], list_[-1])
    return '{0}'.format(list_[0])

# Setup: The function takes an iterable and converts it to a range string
# It calls list() on the iterable, checks len(), and accesses elements [0] and [-1]
# Elements need to be formattable as strings

# Test case 1: Multiple elements (will use the range format)
ARGS = {
    'iterable': [1, 2, 3, 4, 5]
}
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 str — a str output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a str 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
atoms_in_path Ran
ran on a synthesized input
camall3n/focused-macros/experiments/search.py
code served (permissive licence) · get_code("d564444544bc9056")
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 set. 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 numpy as np

# Define a simple Node class that the function expects
class Node:
    def __init__(self, state, parent=None, h_score=0):
        self.state = state
        self.parent = parent
        self.h_score = h_score

# Create a small chain of nodes with states
# States are tuples containing position-value pairs
state1 = (1, 2, 3, 4)
state2 = (1, 2, 5, 4)
state3 = (1, 7, 3, 4)

# Build a linked list of nodes with decreasing h_scores
node3 = Node(state=state3, parent=None, h_score=1)
node2 = Node(state=state2, parent=node3, h_score=2)
node1 = Node(state=state1, parent=node2, h_score=3)

ARGS = {
    "node": node1,
    "any_h": False
}
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 set — a set output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a set 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_face_swaps Ran
ran on a synthesized input
camall3n/skills-for-planning/domains/cube/cube.py
code served (permissive licence) · get_code("346ba73df65ff479")
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 list, ndim 3, dtype family float; 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 enum

class Pos(enum.IntEnum):
    """Enum for representing the position of a sticker on a Face

    Positions are named for their corresponding compass directions, plus M for middle
    """
    NW = 0
    N = 1
    NE = 2
    W = 3
    M = 4
    E = 5
    SW = 6
    S = 7
    SE = 8

def get_face_swaps(face):
    """Return the list of position swaps on a single face associated with rotating that face

    Each swap is a tuple of tuples: ((face, StartPos), (face, EndPos))

    The rotation for a given face is clockwise w.r.t. the 2-D plane in the render diagram
    """
    swaps = [((face, Pos.NW), (face, Pos.NE)), ((face, Pos.N), (face, Pos.E)), ((face, Pos.NE), (face, Pos.SE)), ((face, Pos.E), (face, Pos.S)), ((face, Pos.SE), (face, Pos.SW)), ((face, Pos.S), (face, Pos.W)), ((face, Pos.SW), (face, Pos.NW)), ((face, Pos.W), (face, Pos.N))]
    return swaps

# The function takes a single parameter 'face' which is used as an identifier
# in constructing tuples. It can be any value - string, int, etc.
ARGS = {
    "face": "U"
}
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 list, ndim 3, dtype family float — a list 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
get_inverse_swaps Ran
ran on a synthesized input
camall3n/skills-for-planning/domains/cube/cube.py
code served (permissive licence) · get_code("c6559ef4f33eede6")
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 list, ndim 3, 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

# Construct a swap_list with the expected structure
# Each swap is ((StartFace, StartPos), (EndFace, EndPos))
swap_list = [
    ((0, 1), (2, 3)),
    ((1, 2), (3, 4)),
    ((2, 0), (1, 5))
]

ARGS = {
    "swap_list": swap_list
}
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 list, ndim 3, dtype family int — a list 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
get_unique_atoms Ran
ran on a synthesized input
camall3n/focused-macros/experiments/search.py
code served (permissive licence) · get_code("e12b3dd5023ee1ed")
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 set. 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 numpy as np

torch.manual_seed(0)

# The function takes 'states' parameter which is iterated over.
# Each state is expected to be iterable (can be enumerated with pos, val).
# The states are converted to tuples for hashing/sorting.
# Let's create a list of states where each state is a tuple or list of values.

states = [
    (1, 2, 3, 0),
    (1, 2, 3, 0),
    (0, 2, 3, 1),
    (1, 0, 3, 2),
]

ARGS = {
    "states": states
}
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 set — a set output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a set 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
inverse Ran
ran on a synthesized input
camall3n/skills-for-planning/domains/cube/formula.py
code served (permissive licence) · get_code("0ce89e754ee8249a")
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 list, ndim 1, dtype family float; 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 copy

def _inverse_move(move):
    """Invert a move"""
    if "'" in move:
        return move.strip("'")
    return move + "'"

def inverse(formula):
    """Invert a formula"""
    result = copy.copy(formula)
    result.reverse()
    for i, move in enumerate(result):
        result[i] = _inverse_move(move)
    return result

# Construct valid arguments
# The function takes a formula parameter which is a list of move strings
# It uses copy.copy(), reverse(), and iterates with enumerate and _inverse_move

formula_example = ["R", "U'", "F", "D'", "L"]

ARGS = {
    "formula": formula_example
}
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 list, ndim 1, dtype family float — a list 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
ranges_to_string Ran
ran on a synthesized input
camall3n/focused-macros/experiments/show_missing.py
code served (permissive licence) · get_code("2fde2a93736ab459")
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 str. 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
from itertools import groupby, count

torch.manual_seed(0)

# The function ranges_to_string takes a single parameter: range_
# which should be an iterable of numbers that will be grouped into ranges
# Based on the function body, it expects a list/iterable of integers

ARGS = {
    'range_': [1, 2, 3, 5, 6, 7, 8, 10, 12, 13, 14]
}
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 str — a str output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a str 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
reconstruct_path Ran
ran on a synthesized input
camall3n/focused-macros/experiments/search.py
code served (permissive licence) · get_code("a46b83c0375c56e2")
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. 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 numpy as np

class SearchNode:
    def __init__(self, state, action=None, parent=None):
        self.state = state
        self.action = action
        self.parent = parent

# Build a simple search tree structure
leaf_state = torch.tensor([1.0, 2.0, 3.0, 4.0])
leaf_node = SearchNode(state=leaf_state, action="action_3", parent=None)

middle_state = torch.tensor([0.5, 1.5, 2.5, 3.5])
middle_node = SearchNode(state=middle_state, action="action_2", parent=leaf_node)

root_state = torch.tensor([0.1, 0.2, 0.3, 0.4])
root_node = SearchNode(state=root_state, action="action_1", parent=middle_node)

# Create the node to pass to reconstruct_path
node = root_node

ARGS = {
    "node": node
}
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 — a tuple output has no array shape of its own; none is 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
autoscale_xticks Not yet run
unverified (harvested, no recorded run)
camall3n/focused-macros/experiments/plot_planning_time.py
code served (permissive licence) · get_code("b3debbd1467f3153")
How it ran

no invocation recorded.

get_position_swaps Not yet run
unverified (harvested, no recorded run)
camall3n/skills-for-planning/domains/cube/cube.py
code served (permissive licence) · get_code("affd6b324f19c083")
How it ran

no invocation recorded.

load_data Not yet run
unverified (harvested, no recorded run)
camall3n/focused-macros/experiments/plot_planning_time.py
code served (permissive licence) · get_code("51d88f91c690fd88")
How it ran

no invocation recorded.

load_learned_macros Not yet run
unverified (harvested, no recorded run)
camall3n/focused-macros/domains/pddlgym/macros.py
code served (permissive licence) · get_code("85c4ab18f7c52f19")
How it ran

no invocation recorded.

parse_filepath Not yet run
unverified (harvested, no recorded run)
camall3n/focused-macros/experiments/plot_planning_time.py
code served (permissive licence) · get_code("ccb8967a93c905a8")
How it ran

no invocation recorded.

Also present in other papers' code

Of 13 samples lifted from this paper's linked repositories, 13 carry a function identity and 0 do not. Of the 13 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 13 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

The difficulty of deterministic planning increases exponentially with search-tree depth. Black-box planning presents an even greater challenge, since planners must operate without an explicit model of the domain. Heuristics can make search more efficient, but goal-aware heuristics for black-box planning usually rely on goal counting, which is often quite uninformative. In this work, we show how to overcome this limitation by discovering macro-actions that make the goal-count heuristic more accurate. Our approach searches for macro-actions with focused effects (i.e. macros that modify only a small number of state variables), which align well with the assumptions made by the goal-count heuristic. Focused macros dramatically improve black-box planning efficiency across a wide range of planning domains, sometimes beating even state-of-the-art planners with access to a full domain model.

For agents

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

get_harvested_code_for_paper("2004.13242")
get_code_for_paper("2004.13242")
have("2004.13242")

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

curl https://syntology.ai/api/ran/2004.13242.json

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

[![Syntology run record](https://syntology.ai/api/ran/2004.13242.svg)](https://syntology.ai/paper/2004.13242)

Connect an agent — have() is free.