SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2206.01626 · NeurIPS · 2022

Reincarnating Reinforcement Learning: Reusing Prior Computation to Accelerate Progress

Rishabh Agarwal, Aaron Courville, Marc Bellemare, Max Schwarzer, Pablo Castro

arXiv · PDF · Open in the Atlas

Code that ran

2 of 5 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 2 have no contract verdict; 0 failures of our instrument; 3 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/2206.01626.json · badge: /api/ran/2206.01626.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
google-research/reincarnating_rl canonical (classifier label)
classifier basis not recorded on the link
2 ran · 3 no recorded run
FunctionStatusWhere it lives
preprocess_atari_inputs Ran
ran on a synthesized input
google-research/reincarnating_rl/reincarnating_rl/reincarnation_networks.py
code served (permissive licence) · get_code("922cb4050cda4895")
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 ArrayImpl, ndim 4, dtype family float, shape (2, 4, 4, 1) 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 jax.numpy as jnp

# Create a small mock Atari frame input
# Atari frames are typically uint8 arrays representing pixel values (0-255)
# Create a batch of 2 frames with shape (2, 84, 84, 1) - standard Atari preprocessing
x = jnp.array([
    [[[100], [150], [200], [250]],
     [[50], [75], [125], [175]],
     [[10], [20], [30], [40]],
     [[5], [15], [25], [35]]],
    [[[200], [150], [100], [50]],
     [[175], [125], [75], [25]],
     [[40], [30], [20], [10]],
     [[35], [25], [15], [5]]]
], dtype=jnp.uint8)

ARGS = {
    "x": x
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
jax — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ArrayImpl, ndim 4, dtype family float, shape (2, 4, 4, 1) — shape (2, 4, 4, 1) 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
select_roll_out_action Ran
ran on a synthesized input
google-research/reincarnating_rl/reincarnating_rl/jsrl_dqn_agent.py
code served (permissive licence) · get_code("d1504739b277fc57")
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 jax
import jax.numpy as jnp
import jax.random

# Define a decay function that takes (decay_period, training_steps, arg3, arg4)
def decay_fn(decay_period, training_steps, arg3, arg4):
    """Simple decay function that returns a scalar between 0 and 1."""
    # Simple linear decay: starts at 1.0 and decays to 0.0
    decay_rate = jnp.maximum(0.0, 1.0 - (training_steps / jnp.maximum(decay_period, 1.0)))
    return decay_rate

# Create RNG key
rng = jax.random.PRNGKey(0)

# Define arguments
ARGS = {
    "roll_out_guide_prob": 0.8,  # scalar probability
    "roll_out_decay_fn": decay_fn,  # function that returns decay multiplier
    "roll_out_decay_period": 100,  # decay period
    "training_steps": 50,  # current training step
    "rng": rng,  # JAX random key
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
jax — 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
create_dqn_checkpoint_data Not yet run
unverified (harvested, no recorded run)
google-research/reincarnating_rl/reincarnating_rl/checkpoint_helpers.py
code served (permissive licence) · get_code("fc1817cb2b9acbcf")
How it ran

no invocation recorded.

get_logits_and_q_values Not yet run
unverified (harvested, no recorded run)
google-research/reincarnating_rl/reincarnating_rl/qdagger_rainbow_agent.py
code served (permissive licence) · get_code("e7df8658108f2d6b")
How it ran

no invocation recorded.

load_tf_nature_dqn_weights Not yet run
unverified (harvested, no recorded run)
google-research/reincarnating_rl/reincarnating_rl/checkpoint_helpers.py
code served (permissive licence) · get_code("a4f5fad305ee0590")
How it ran

no invocation recorded.

Also present in other papers' code

Of 5 samples lifted from this paper's linked repositories, 5 carry a function identity and 0 do not. Of the 5 distinct functions with an identity, 1 are also present in other papers whose code we hold (1 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.

FunctionAlso present inEarliest-datedOthers named
preprocess_atari_inputs 1 other paper
1 dated earlier
2112.05848 (2021-12)
Faster Deep Reinforcement Learning with Slower Online Network Kavosh Asadi Amazon Web Services

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 tabula rasa, that is without any previously learned knowledge, is the prevalent workflow in reinforcement learning (RL) research. However, RL systems, when applied to large-scale settings, rarely operate tabula rasa. Such large-scale systems undergo multiple design or algorithmic changes during their development cycle and use ad hoc approaches for incorporating these changes without re-training from scratch, which would have been prohibitively expensive. Additionally, the inefficiency of deep RL typically excludes researchers without access to industrialscale resources from tackling computationally-demanding problems. To address these issues, we present reincarnating RL as an alternative workflow or class of problem settings, where prior computational work (e.g., learned policies) is reused or transferred between design iterations of an RL agent, or from one RL agent to another. As a step towards enabling reincarnating RL from any agent to any other agent, we focus on the specific setting of efficiently transferring an existing sub-optimal policy to a standalone value-based RL agent. We find that existing approaches fail in this setting and propose a simple algorithm to address their limitations. Equipped with this algorithm, we demonstrate reincarnating RL's gains over tabula rasa RL on Atari 2600 games, a challenging locomotion task, and the real-world problem of navigating stratospheric balloons. Overall, this work argues for an alternative approach to RL research, which we believe could significantly improve real-world RL adoption and help democratize it further. Open-sourced code and trained agents at agarwl.github.io/reincarnating_rl.

For agents

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

get_harvested_code_for_paper("2206.01626")
get_code_for_paper("2206.01626")
have("2206.01626")

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

curl https://syntology.ai/api/ran/2206.01626.json

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

[![Syntology run record](https://syntology.ai/api/ran/2206.01626.svg)](https://syntology.ai/paper/2206.01626)

Connect an agent — have() is free.