SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2301.12876 · 2023

Guiding Online Reinforcement Learning with Action-Free Offline Pretraining

arXiv · PDF · Open in the Atlas

Code that ran

1 of 1 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, 0 violated it and 0 have no contract verdict; 0 failures of our instrument; 0 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/2301.12876.json · badge: /api/ran/2301.12876.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
vision-cair/af-guide canonical (classifier label)
classifier basis: classifier_2025
1 ran · 0 no recorded run
FunctionStatusWhere it lives
eval_valid Ran
ran, honoured the contract we drafted
vision-cair/af-guide/scripts/train_afdt.py
code served (permissive licence) · get_code("0e963f3dff9681ab")
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 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 torch
import torch.nn as nn
import numpy as np

torch.manual_seed(0)

# Define a minimal model that satisfies the function's expectations
class MinimalModel(nn.Module):
    def __init__(self, state_dim=4, act_dim=3):
        super().__init__()
        self.state_dim = state_dim
        self.act_dim = act_dim
        self.linear = nn.Linear(state_dim + act_dim + 1, state_dim + act_dim + 1)
    
    def forward(self, timesteps, states, actions, returns_to_go):
        # timesteps: (batch, seq_len)
        # states: (batch, seq_len, state_dim)
        # actions: (batch, seq_len, act_dim)
        # returns_to_go: (batch, seq_len, 1)
        batch_size, seq_len = states.shape[0], states.shape[1]
        
        # Return dummy predictions with correct shapes
        diff_state_preds = torch.randn(batch_size, seq_len, self.state_dim)
        action_preds = torch.randn(batch_size, seq_len, self.act_dim)
        return_preds = torch.randn(batch_size, seq_len, 1)
        
        return diff_state_preds, action_preds, return_preds

# Create model
model = MinimalModel(state_dim=4, act_dim=3)

# Create a minimal data iterator (list of batches)
# Each batch contains: (timesteps, states, actions, returns_to_go, traj_mask)
batch_size = 2
seq_len = 5
state_dim = 4
act_dim = 3

timesteps = torch.randint(0, 10, (batch_size, seq_len))
states = torch.randn(batch_size, seq_len, state_dim)
actions = torch.randn(batch_size, seq_len, act_dim)
returns_to_go = torch.randn(batch_size, seq_len)
traj_mask = torch.ones(batch_size, seq_len)

data_iter = [
    (timesteps, states, actions, returns_to_go, traj_mask),
    (timesteps, states, actions, returns_to_go, traj_mask)
]

# Scalars for normalization
state_std = torch.tensor(1.5)
diff_state_std = torch.tensor(0.8)
diff_state_mean = torch.tensor(0.2)

device = 'cpu'

ARGS = {
    'model': model,
    'data_iter': data_iter,
    'device': device,
    'state_std': state_std,
    'diff_state_std': diff_state_std,
    'diff_state_mean': diff_state_mean
}
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, ndim 1, dtype family float — 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

Also present in other papers' code

Of 1 samples lifted from this paper's linked repositories, 1 carry a function identity and 0 do not. Of the 1 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 1 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

Offline RL methods have been shown to reduce the need for environment interaction by training agents using offline collected episodes. However, these methods typically require action information to be logged during data collection, which can be difficult or even impossible in some practical cases. In this paper, we investigate the potential of using action-free offline datasets to improve online reinforcement learning, name this problem Reinforcement Learning with Action-Free Offline Pretraining (AFP-RL). We introduce Action-Free Guide (AF-Guide), a method that guides online training by extracting knowledge from action-free offline datasets. AF-Guide consists of an Action-Free Decision Transformer (AFDT) implementing a variant of Upside-Down Reinforcement Learning. It learns to plan the next states from the offline dataset, and a Guided Soft Actor-Critic (Guided SAC) that learns online with guidance from AFDT. Experimental results show that AF-Guide can improve sample efficiency and performance in online training thanks to the knowledge from the action-free offline dataset. Code is available at https://github.com/Vision-CAIR/AF-Guide.

For agents

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

get_harvested_code_for_paper("2301.12876")
get_code_for_paper("2301.12876")
have("2301.12876")

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

curl https://syntology.ai/api/ran/2301.12876.json

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

[![Syntology run record](https://syntology.ai/api/ran/2301.12876.svg)](https://syntology.ai/paper/2301.12876)

Connect an agent — have() is free.