SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1805.09613 · 2018

A0C: Alpha Zero in Continuous Action Space

arXiv · PDF · Open in the Atlas

Code that ran

3 of 3 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, 1 violated it and 0 have no contract verdict; 2 failures of our instrument; 0 no recorded run), measured 2026-11-21; 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/1805.09613.json · badge: /api/ran/1805.09613.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
tmoer/a0c reimplementation (classifier label)
classifier basis: classifier_2025
3 ran · 0 no recorded run
FunctionStatusWhere it lives
bring_Env_to_correct_state Ran
ran; our contract draft was wrong, not the code
tmoer/a0c/src/mcts.py
code served (permissive licence) · get_code("bc47c92dab08c505")
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 MockEnv. 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

# Create a minimal mock environment that satisfies the function's needs
class MockEnv:
    def __init__(self):
        self.state = np.array([0, 1, 2, 3])
        self.action_space = type('obj', (object,), {'n': 4})()
    
    def reset(self):
        self.state = np.array([0, 1, 2, 3])
        return self.state
    
    def step(self, action):
        # Simulate a step: modify state based on action
        self.state = self.state + action
        reward = float(action)
        done = False
        info = {}
        return self.state, reward, done, info

# Construct arguments
Env = MockEnv()
seed = 42
a_his = [1, 2, 0, 1]

ARGS = {
    "Env": Env,
    "seed": seed,
    "a_his": a_his
}
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 MockEnv — a MockEnv output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a MockEnv 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
check_for_loop_in_trace Ran
ran, violated the contract we drafted
tmoer/a0c/src/mcts.py
code served (permissive licence) · get_code("e4223b15b42ab2b0")
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 bool, ndim 0, dtype family float, shape () 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

class MockAction:
    def __init__(self, parent_state=None):
        self.parent_state = parent_state

class MockState:
    def __init__(self, index, parent_action=None):
        self.index = index
        self.parent_action = parent_action

# Create a chain of states with actions linking them
# State 0 -> Action 0 -> State 1 -> Action 1 -> State 2 -> Action 2 -> State 0 (loop back)
state_0 = MockState(index=np.array([0.0, 1.0, 2.0, 3.0]))
action_0 = MockAction(parent_state=None)
state_0.parent_action = None

state_1 = MockState(index=np.array([0.5, 1.5, 2.5, 3.5]))
action_1 = MockAction(parent_state=state_0)
state_1.parent_action = action_1

state_2 = MockState(index=np.array([0.001, 1.001, 2.001, 3.001]))  # Very close to state_0
action_2 = MockAction(parent_state=state_1)
state_2.parent_action = action_2
action_1.parent_state = state_2

# Now state_0's parent_action points to action_2, which points back through the chain
state_0.parent_action = action_2

ARGS = {
    "state": state_2,
    "threshold": 0.01
}
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 bool, ndim 0, dtype family float, shape () — shape () 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
override_hps_settings Ran
ran; our contract draft was wrong, not the code
tmoer/a0c/config/hps.py
code served (permissive licence) · get_code("2c57a27d115bfe60")
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 HPS. 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 HPS:
    def __init__(self):
        self.mode = 'off'

hps = HPS()

ARGS = {
    'hps': hps
}
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 HPS — a HPS output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a HPS 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

Also present in other papers' code

Of 3 samples lifted from this paper's linked repositories, 3 carry a function identity and 0 do not. Of the 3 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 3 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 2026-11-21.

Repositories linked to this paper

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

Abstract

A core novelty of Alpha Zero is the interleaving of tree search and deep learning, which has proven very successful in board games like Chess, Shogi and Go. These games have a discrete action space. However, many real-world reinforcement learning domains have continuous action spaces, for example in robotic control, navigation and self-driving cars. This paper presents the necessary theoretical extensions of Alpha Zero to deal with continuous action space. We also provide some preliminary experiments on the Pendulum swing-up task, empirically showing the feasibility of our approach. Thereby, this work provides a first step towards the application of iterated search and learning in domains with a continuous action space.

For agents

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

get_harvested_code_for_paper("1805.09613")
get_code_for_paper("1805.09613")
have("1805.09613")

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

curl https://syntology.ai/api/ran/1805.09613.json

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

[![Syntology run record](https://syntology.ai/api/ran/1805.09613.svg)](https://syntology.ai/paper/1805.09613)

Connect an agent — have() is free.