SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2307.04738 · 2023

RoCo: Dialectic Multi-Robot Collaboration with Large Language Models

arXiv · PDF · Open in the Atlas

Code that ran

6 of 8 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; 1 honoured the contract we drafted, 1 violated it and 0 have no contract verdict; 4 failures of our instrument; 2 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/2307.04738.json · badge: /api/ran/2307.04738.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
MandiZhao/robot-collab canonical (classifier label)
classifier basis: census
6 ran · 2 no recorded run
FunctionStatusWhere it lives
compose_html_block Ran
ran; our contract draft was wrong, not the code
MandiZhao/robot-collab/prompting/display_utils.py
code served (permissive licence) · get_code("a55775e13718dc34")
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 torch.nn as nn

# Analyze the function: compose_html_block
# Parameters:
# - container_name: str (default 'chat-container')
# - include_video: bool (default True)
# - video_fname: str (default 'execute.mp4')
# 
# The function uses these parameters in string formatting and conditionals.
# No complex objects needed - all parameters are primitives (str, bool).

ARGS = {
    'container_name': 'chat-container',
    'include_video': True,
    'video_fname': 'execute.mp4'
}
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
compose_html_file Ran
ran; our contract draft was wrong, not the code
MandiZhao/robot-collab/prompting/display_utils.py
code served (permissive licence) · get_code("b345b7419040e3c3")
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

ARGS = {
    'js_fname': 'script.js',
    'css_code': 'body { font-family: Arial, sans-serif; } .message { padding: 10px; margin: 5px 0; border: 1px solid #ccc; }',
    'html_blocks': [
        '<div class="message"><strong>User:</strong> Hello</div>',
        '<div class="message"><strong>Bot:</strong> Hi there!</div>',
        '<div class="message"><strong>User:</strong> How are you?</div>'
    ]
}
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
compose_js_code Ran
ran; our contract draft was wrong, not the code
MandiZhao/robot-collab/prompting/display_utils.py
code served (permissive licence) · get_code("fe7ced4cbc166722")
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 json

# Construct valid arguments for compose_js_code
# The function takes container_idx, data, json_fname, and include_function

# Create sample chat data
chat_data = [
    {"sender": "Alice", "message": "Hello, how are you?"},
    {"sender": "Bob", "message": "I'm doing great, thanks for asking!"},
    {"sender": "Alice", "message": "That's wonderful to hear."}
]

ARGS = {
    "container_idx": 0,
    "data": chat_data,
    "json_fname": None,
    "include_function": True
}
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 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
get_workspace_crop Ran
ran, violated the contract we drafted
MandiZhao/robot-collab/real_world/kinect.py
code served (permissive licence) · get_code("6bebdc06ccbba5ee")
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 Tensor, ndim 4, dtype family float, shape (0, 0, 3, 2) 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 torch
import numpy as np

torch.manual_seed(0)

# The function get_workspace_crop takes an img parameter and slices it using indices
# WS_PC = [180, -120, 300, -380]
# The slicing is: img[180:-120, 300:-380, ...]
# This means img needs to be at least a 2D or higher dimensional array

# Create a small image tensor that satisfies the slicing requirements
# Need at least 181 rows (for index 180) and 301 columns (for index 300)
# Using distinct sizes: 200 rows, 400 columns, 3 channels, 2 batch
img = torch.randn(200, 400, 3, 2)

ARGS = {
    "img": img
}
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 Tensor, ndim 4, dtype family float, shape (0, 0, 3, 2) — shape (0, 0, 3, 2) 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
setup_thread Ran
ran; our contract draft was wrong, not the code
MandiZhao/robot-collab/real_world/realur5_utils.py
code served (permissive licence) · get_code("7aa04c0a400f223c")
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 Thread. 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 threading
from threading import Thread

def simple_target():
    """A simple callable that can be used as the target for a thread."""
    pass

ARGS = {
    'target': simple_target
}
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 Thread — a Thread output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Thread 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
skip_to_package_index Ran
ran, honoured the contract we drafted
MandiZhao/robot-collab/real_world/realur5_utils.py
code served (permissive licence) · get_code("a7dcb50288ee32d5")
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 int, ndim 0, dtype family int, 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 struct

def skip_to_package_index(state_data, pkg_type):
    byte_index = 1
    while byte_index < len(state_data):
        package_size = struct.unpack('!i', state_data[byte_index:byte_index + 4])[0]
        byte_index += 4
        package_index = int(struct.unpack('!B', state_data[byte_index + 0:byte_index + 1])[0])
        if package_index == pkg_type:
            byte_index += 1
            break
        byte_index += package_size - 4
    return byte_index

# Construct state_data as a binary blob with package structures
# Format: 1 byte initial offset, then repeated: 4 bytes size (big-endian int) + 1 byte index
# We need at least one package with matching pkg_type

state_data = bytearray()
state_data.append(0)  # Initial offset byte

# Package 1: size=10, index=1 (not matching)
state_data.extend(struct.pack('!i', 10))
state_data.append(1)
state_data.extend(b'\x00' * 5)  # padding to reach size 10

# Package 2: size=8, index=42 (this will match pkg_type=42)
state_data.extend(struct.pack('!i', 8))
state_data.append(42)
state_data.extend(b'\x00' * 3)  # padding to reach size 8

state_data = bytes(state_data)

ARGS = {
    'state_data': state_data,
    'pkg_type': 42
}
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 int, ndim 0, dtype family int, 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
calibrate Not yet run
unverified (harvested, no recorded run)
MandiZhao/robot-collab/real_world/calibration_robot.py
code served (permissive licence) · get_code("0083cd2cd42d2108")
How it ran

no invocation recorded.

connect Not yet run
unverified (harvested, no recorded run)
MandiZhao/robot-collab/real_world/realur5_utils.py
code served (permissive licence) · get_code("91f5592ce8fe7689")
How it ran

no invocation recorded.

Also present in other papers' code

Of 8 samples lifted from this paper's linked repositories, 8 carry a function identity and 0 do not. Of the 8 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 8 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

We propose a novel approach to multi-robot collaboration that harnesses the power of pre-trained large language models (LLMs) for both high-level communication and low-level path planning. Robots are equipped with LLMs to discuss and collectively reason task strategies. They then generate sub-task plans and task space waypoint paths, which are used by a multi-arm motion planner to accelerate trajectory planning. We also provide feedback from the environment, such as collision checking, and prompt the LLM agents to improve their plan and waypoints in-context. For evaluation, we introduce RoCoBench, a 6-task benchmark covering a wide range of multi-robot collaboration scenarios, accompanied by a text-only dataset for agent representation and reasoning. We experimentally demonstrate the effectiveness of our approach -- it achieves high success rates across all tasks in RoCoBench and adapts to variations in task semantics. Our dialog setup offers high interpretability and flexibility -- in real world experiments, we show RoCo easily incorporates human-in-the-loop, where a user can communicate and collaborate with a robot agent to complete tasks together. See project website https://project-roco.github.io for videos and code.

For agents

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

get_harvested_code_for_paper("2307.04738")
get_code_for_paper("2307.04738")
have("2307.04738")

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

curl https://syntology.ai/api/ran/2307.04738.json

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

[![Syntology run record](https://syntology.ai/api/ran/2307.04738.svg)](https://syntology.ai/paper/2307.04738)

Connect an agent — have() is free.