SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1901.08129 · 2019

The Multi-Agent Reinforcement Learning in MalmÖ (MARLÖ) Competition

arXiv · PDF · Open in the Atlas

Code that ran

3 of 4 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 3 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/1901.08129.json · badge: /api/ran/1901.08129.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
crowdAI/marLo community (classifier label)
classifier basis not recorded on the link
3 ran · 1 no recorded run
FunctionStatusWhere it lives
eval_performance Ran
ran on a synthesized input
crowdAI/marLo/marlo/experiments/evaluator.py
code served (permissive licence) · get_code("9888ad094311172f")
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 logging
import numpy as np

# Create a minimal Environment class
class DummyEnv:
    def __init__(self):
        self.observation_space_shape = (4,)
        self.action_space_n = 2
    
    def reset(self):
        return np.array([0.1, 0.2, 0.3, 0.4], dtype=np.float32)
    
    def step(self, action):
        # Returns: obs, reward, done, info
        obs = np.array([0.1, 0.2, 0.3, 0.4], dtype=np.float32)
        reward = float(np.random.rand())
        done = False
        info = {}
        return obs, reward, done, info

# Create a minimal Agent class
class DummyAgent:
    def __init__(self):
        self.episode_count = 0
    
    def act(self, obs):
        # Return a simple action (integer)
        return int(np.random.randint(0, 2))
    
    def stop_episode(self):
        self.episode_count += 1

# Create a minimal Explorer class (optional)
class DummyExplorer:
    def select_action(self, t, greedy_action_func):
        # With some probability, use greedy action
        if np.random.rand() > 0.1:
            return greedy_action_func()
        else:
            return int(np.random.randint(0, 2))

# Set up logging
logger = logging.getLogger(__name__)
logger.setLevel(logging.INFO)
handler = logging.StreamHandler()
logger.addHandler(handler)

# Create instances
env = DummyEnv()
agent = DummyAgent()
n_runs = 2
max_episode_len = 10
explorer = DummyExplorer()

ARGS = {
    'env': env,
    'agent': agent,
    'n_runs': n_runs,
    'max_episode_len': max_episode_len,
    'explorer': explorer,
    'logger': logger
}
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
run_evaluation_episodes Ran
ran on a synthesized input
crowdAI/marLo/marlo/experiments/evaluator.py
code served (permissive licence) · get_code("b0a9e1e280bb73f0")
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 logging
import torch
import torch.nn as nn

torch.manual_seed(0)

# Mock Environment class
class MockEnv:
    def __init__(self):
        self.obs_shape = (4,)
        self.action_space_size = 3
    
    def reset(self):
        return torch.randn(4)
    
    def step(self, action):
        obs = torch.randn(4)
        reward = 1.0
        done = False
        info = {}
        return obs, reward, done, info

# Mock Agent class
class MockAgent(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear = nn.Linear(4, 3)
    
    def act(self, obs):
        if isinstance(obs, torch.Tensor):
            obs_tensor = obs
        else:
            obs_tensor = torch.tensor(obs, dtype=torch.float32)
        if obs_tensor.dim() == 1:
            obs_tensor = obs_tensor.unsqueeze(0)
        action = self.linear(obs_tensor).argmax(dim=1).item()
        return action
    
    def stop_episode(self):
        pass

# Mock Explorer class
class MockExplorer:
    def select_action(self, timestep, greedy_action_func):
        return greedy_action_func()

# Create instances
env = MockEnv()
agent = MockAgent()
n_runs = 2
max_episode_len = 10
explorer = MockExplorer()
logger = logging.getLogger(__name__)

ARGS = {
    'env': env,
    'agent': agent,
    'n_runs': n_runs,
    'max_episode_len': max_episode_len,
    'explorer': explorer,
    'logger': logger
}
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 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
threaded Ran
ran on a synthesized input
crowdAI/marLo/marlo/utils.py
code served (permissive licence) · get_code("d5bf0d80fd025e0a")
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 function. 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 time
from threading import Thread
from queue import Queue

class ExceptionHolder:
    def __init__(self, exn):
        self.exception = exn

def threaded(fn):
    """Implements the ``@marlo.threaded`` decorator to help easily run functions in a 
    separate thread. Useful in multiagent scenarios when we want to run 
    multiple blocking agents across different threads.

    .. code-block:: python

        import marlo

        @marlo.threaded
        def example_function(agent_id):
            print("Agent-id : {}; sleeping for two seconds".format(agent_id))
            time.sleep(2)
            print("Exiting : {} ".format(agent_id))

        thread_handler_1, _ = example_function(1)
        thread_handler_2, _ = example_function(2)

        thread_handler_1.join()
        thread_handler_2.join()
        print("Code Exit")

    :param fn: Function defitnion 
    :type fn: func
    
    :returns thread_handler to join the threads if required
    """

    def wrap(queue, *args, **kwargs):
        try:
            queue.put(fn(*args, **kwargs))
        except Exception as e:
            print('Exception in threaded function: ' + str(e))
            queue.put(ExceptionHolder(e))

    def call(*args, **kwargs):
        queue = Queue()
        job = Thread(target=wrap, args=(queue,) + args, kwargs=kwargs)
        job.start()
        return (job, queue)
    return call


def sample_function(agent_id, delay=0.1):
    """A simple sample function to decorate with @threaded"""
    time.sleep(delay)
    return f"Agent {agent_id} completed"


ARGS = {
    'fn': sample_function
}
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 function — a function output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a function 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
launch_minecraft_in_background Not yet run
unverified (harvested, no recorded run)
crowdAI/marLo/marlo/launch_minecraft_in_background.py
code served (permissive licence) · get_code("0cdcdcf9fe7fe29d")
How it ran

no invocation recorded.

Also present in other papers' code

Of 4 samples lifted from this paper's linked repositories, 4 carry a function identity and 0 do not. Of the 4 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 4 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

Learning in multi-agent scenarios is a fruitful research direction, but current approaches still show scalability problems in multiple games with general reward settings and different opponent types. The Multi-Agent Reinforcement Learning in MalmÖ (MARLÖ) competition is a new challenge that proposes research in this domain using multiple 3D games. The goal of this contest is to foster research in general agents that can learn across different games and opponent types, proposing a challenge as a milestone in the direction of Artificial General Intelligence.

For agents

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

get_harvested_code_for_paper("1901.08129")
get_code_for_paper("1901.08129")
have("1901.08129")

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

curl https://syntology.ai/api/ran/1901.08129.json

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

[![Syntology run record](https://syntology.ai/api/ran/1901.08129.svg)](https://syntology.ai/paper/1901.08129)

Connect an agent — have() is free.