SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2112.10751 · 2021

RvS: What is Essential for Offline RL via Supervised Learning?

arXiv · PDF · Open in the Atlas

Code that ran

8 of 9 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; 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/2112.10751.json · badge: /api/ran/2112.10751.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
scottemmons/rvs canonical (classifier label)
classifier basis not recorded on the link
8 ran · 1 no recorded run
FunctionStatusWhere it lives
configure_gpu Ran
ran on a synthesized input
scottemmons/rvs/src/rvs/util.py
code served (permissive licence) · get_code("904308085f6cca90")
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 device. The image it ran in was not recorded. The code mentions cuda, and the harness aliases cuda to the CPU for every sample; whether this run needed that is 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 os
import torch

torch.manual_seed(0)

ARGS = {
    'use_gpu': False,
    'which_gpu': 0,
}
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 device — a device output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a device 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
extract_done_markers Ran
ran on a synthesized input
scottemmons/rvs/src/rvs/util.py
code served (permissive licence) · get_code("8b1664741c54346a")
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 2, 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 numpy as np

# Create a simple dones vector that marks episode terminations
# The function extracts where episodes end (True values) and computes trajectory segments
dones = np.array([False, False, True, False, False, True, False], dtype=bool)

ARGS = {
    "dones": dones
}
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 tuple, ndim 2, 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
extract_traj_markers Ran
ran on a synthesized input
scottemmons/rvs/src/rvs/util.py
code served (permissive licence) · get_code("0bab79bd43ebf743")
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 2, 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 numpy as np

# Create a minimal D4RL-style dataset with terminals and timeouts
dataset = {
    'terminals': np.array([False, False, True, False, False, True, False]),
    'timeouts': np.array([False, False, False, False, True, False, False]),
}

ARGS = {
    'dataset': dataset,
}
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 tuple, ndim 2, 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
get_total_steps Ran
ran on a synthesized input
scottemmons/rvs/src/rvs/step.py
code served (permissive licence) · get_code("6aefaad3876d1cda")
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 os
import tempfile
import numpy as np

# Create a temporary directory and populate it with the required .npy files
temp_dir = tempfile.mkdtemp()

# Create sample numpy arrays with consistent shapes
# s_obs_vecs: (num_trajectories, steps_per_traj, obs_dim)
s_obs_vecs = np.random.randn(2, 3, 4).astype(np.float32)

# s_ach_goal_vecs: (num_trajectories, steps_per_traj, goal_dim)
s_ach_goal_vecs = np.random.randn(2, 3, 5).astype(np.float32)

# a_vecs: (num_trajectories, steps_per_traj, action_dim)
a_vecs = np.random.randn(2, 3, 6).astype(np.float32)

# Save the arrays to the temporary directory
np.save(os.path.join(temp_dir, 's_obs_vecs.npy'), s_obs_vecs)
np.save(os.path.join(temp_dir, 's_ach_goal_vecs.npy'), s_ach_goal_vecs)
np.save(os.path.join(temp_dir, 'a_vecs.npy'), a_vecs)

ARGS = {
    'rollout_dir': temp_dir
}
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 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
is_kitchen_env Ran
ran on a synthesized input
scottemmons/rvs/src/rvs/step.py
code served (permissive licence) · get_code("092e8da61aa4b415")
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 gym
import numpy as np

# Create a mock environment with spec.id that matches a kitchen env
class MockKitchenEnv(gym.Env):
    def __init__(self, env_id):
        self.spec = type('Spec', (), {'id': env_id})()

# Test with a kitchen environment
env = MockKitchenEnv('kitchen-complete-v0')

ARGS = {
    'env': env
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
gym, 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
load_rollouts Ran
ran on a synthesized input
scottemmons/rvs/src/rvs/step.py
code served (permissive licence) · get_code("408460ecefee36fd")
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 os
import numpy as np
import tempfile
import shutil

# Create a temporary directory for the rollout files
temp_dir = tempfile.mkdtemp()

# Create sample numpy arrays with distinct shapes
s_obs_vecs = np.random.randn(10, 4).astype(np.float32)
s_ach_goal_vecs = np.random.randn(10, 5).astype(np.float32)
a_vecs = np.random.randn(10, 3).astype(np.float32)

# Save the arrays to the temporary directory
np.save(os.path.join(temp_dir, 's_obs_vecs.npy'), s_obs_vecs)
np.save(os.path.join(temp_dir, 's_ach_goal_vecs.npy'), s_ach_goal_vecs)
np.save(os.path.join(temp_dir, 'a_vecs.npy'), a_vecs)

ARGS = {
    'rollout_dir': temp_dir
}
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 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
make_s_g_tensor Ran
ran on a synthesized input
scottemmons/rvs/src/rvs/dataset.py
code served (permissive licence) · get_code("864394f1eb93272b")
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 2, dtype family float, shape (2, 7) 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

torch.manual_seed(0)

# Create small distinct tensors for states and goals
# states: batch_size=2, state_dim=4
states = torch.randn(2, 4)

# goals: batch_size=2, goal_dim=3
goals = torch.randn(2, 3)

ARGS = {
    "states": states,
    "goals": goals,
}
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 Tensor, ndim 2, dtype family float, shape (2, 7) — shape (2, 7) 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
to_tensor_dataset Ran
ran on a synthesized input
scottemmons/rvs/src/rvs/dataset.py
code served (permissive licence) · get_code("459c8b306f2c2123")
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 TensorDataset. 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
from torch.utils import data

torch.manual_seed(0)

# Create sample data: list of tuples (state, goal, action)
# Each state is a numpy array of shape (4,)
# Each goal is a numpy array of shape (3,)
# Each action is a numpy array of shape (2,)

data_vec = [
    (np.array([1.0, 2.0, 3.0, 4.0]), np.array([5.0, 6.0, 7.0]), np.array([0.1, 0.2])),
    (np.array([1.5, 2.5, 3.5, 4.5]), np.array([5.5, 6.5, 7.5]), np.array([0.3, 0.4])),
]

ARGS = {
    "data_vec": data_vec
}
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 TensorDataset — a TensorDataset 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
aggregate_performance Not yet run
unverified (harvested, no recorded run)
scottemmons/rvs/src/rvs/visualize.py
code served (permissive licence) · get_code("e2168f17c792e370")
How it ran

no invocation recorded.

Also present in other papers' code

Of 9 samples lifted from this paper's linked repositories, 9 carry a function identity and 0 do not. Of the 9 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 9 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

Recent work has shown that supervised learning alone, without temporal difference (TD) learning, can be remarkably effective for offline RL. When does this hold true, and which algorithmic components are necessary? Through extensive experiments, we boil supervised learning for offline RL down to its essential elements. In every environment suite we consider, simply maximizing likelihood with a two-layer feedforward MLP is competitive with state-of-the-art results of substantially more complex methods based on TD learning or sequence modeling with Transformers. Carefully choosing model capacity (e.g., via regularization or architecture) and choosing which information to condition on (e.g., goals or rewards) are critical for performance. These insights serve as a field guide for practitioners doing Reinforcement Learning via Supervised Learning (which we coin "RvS learning"). They also probe the limits of existing RvS methods, which are comparatively weak on random data, and suggest a number of open problems.

For agents

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

get_harvested_code_for_paper("2112.10751")
get_code_for_paper("2112.10751")
have("2112.10751")

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

curl https://syntology.ai/api/ran/2112.10751.json

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

[![Syntology run record](https://syntology.ai/api/ran/2112.10751.svg)](https://syntology.ai/paper/2112.10751)

Connect an agent — have() is free.