Papers › Off-Dynamics Reinforcement Learning: Training for Transfer with Domain Classifiers

Off-Dynamics Reinforcement Learning: Training for Transfer with Domain Classifiers

24 Jun 2020ICLR 2021 1arXiv:2006.13916archive 2025-07-28

Benjamin Eysenbach, Swapnil Asawa, Shreyas Chaudhari, Sergey Levine, Ruslan Salakhutdinov

We propose a simple, practical, and intuitive approach for domain adaptation in reinforcement learning. Our approach stems from the idea that the agent's experience in the source domain should look similar to its experience in the target domain. Building off of a probabilistic view of RL, we formally show that we can achieve this goal by compensating for the difference in dynamics by modifying the reward function. This modified reward function is simple to estimate by learning auxiliary classifiers that distinguish source-domain transitions from target-domain transitions. Intuitively, the modified reward function penalizes the agent for visiting states and taking actions in the source domain which are not possible in the target domain. Said another way, the agent is penalized for transitions that would indicate that the agent is interacting with the source domain, rather than the target domain. Our approach is applicable to domains with continuous states and actions and does not require learning an explicit model of the dynamics. On discrete and continuous control tasks, we illustrate the mechanics of our approach and demonstrate its scalability to high-dimensional tasks.

PaperPDFConference PDFCodeCode Syntology ranMethod as code

In Syntology View this paper on Syntology, its page in Syntology's graph. That page adds which of its functions are also present in other papers' code Syntology holds, with dates, and its dated run record as JSON and a README badge. It also lists the repositories linked to the paper, the abstract and the calls for agents.

Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Syntology holds a reference implementation of Off-Dynamics Reinforcement Learning, generated from the paper text (not the authors' code), at level 3 on Syntology's ladder: V2 “two implementations GENERATED from the paper's signature and spec agree numerically (1e-6) on fixture inputs” “neither of the two ever read the authors' own code, so V2 measures how unambiguously the paper specifies the method, not fidelity to the authors' implementation (measured 2026-09-11, commit 75eaa77: cross_checked_by is a second LLM on 2,628 of 2,831 served samples)”; V3 “paper-derived property tests pass on both of those same two generations.” (quoted from Syntology's server card). No rung compares this code against the authors' implementation. For agents: get_reference_implementation(name='Off-Dynamics Reinforcement Learning') (how to connect); the free check have('2006.13916') needs no token.

For agents, on Syntology's MCP service (how to connect):

Code

Syntology 23 code samples harvested from 4 repositories linked to this paper: 21 ran (of which 4 constructed an object rather than computing a result; 14 with no instrument failure: 5 honoured, 1 violated, 8 with no contract checked; 7 where Syntology's instrument failed) · 2 unverified. Where Syntology's instrument failed: 7 where our contract draft was wrong; failures of Syntology's instrument, not of the code.

By repository: community (archive-listed): 17 samples from 2 repositories, 15 ran (10 with no instrument failure; 5 where Syntology's instrument failed); found in paper text by Syntology: 6 samples from 2 repositories, 6 ran (4 with no instrument failure; 2 where Syntology's instrument failed). The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

No sample is linked to this paper from 1 of its repositories that Syntology harvested for other papers:

Levels written between 2026-09-22 and 2026-09-25 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed. 1 of the 21 samples that ran has no recorded date.

pauldaoudi/food mentioned on GitHubpytorchSyntology: 11 samples, 9 ran (4 instrument-failed) · 2 unverified report
shreyasc-13/off_domain_rlkit mentioned on GitHubpytorchMITSyntology: 6 samples, 6 ran (1 instrument-failed) · 0 unverified report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

A paper named beside a repository with no sample linked to this paper is shown with that paper's own run record, not this paper's: “ran” means executed on a synthesized input, not that the code is correct, and “where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Papers are listed in arXiv-id order, at most three per repository.

Code Syntology ran Syntology

23 samples harvested: 21 ran (of which 4 constructed an object rather than computing a result; 14 with no instrument failure: 5 honoured, 1 violated, 8 with no contract checked; 7 where Syntology's instrument failed) · 2 unverified. “Where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Syntology's record for this page last changed ; when this build read Syntology's graph is in the build record. That date is not when the samples ran.

5ran · honoured contract
1ran · violated contract
7ran · our draft was wrong
8ran
2unverified

Licence: 7 of the 23 samples are pointer only, meaning Syntology does not serve that copy's text. For 7 of them, this paper's copy is under a permissive licence and cleared, but the stored copy of the same code that Syntology serves is not; the licence label's tooltip gives Syntology's recorded reason. This page shows no code text for any sample; each one links to its file in the repository.

Harvested from 4 repositories linked to this paper, official or community; the list below is grouped by repository, each group headed by its name and what it is to the paper. “Ran” means the sample executed on a synthesized input. It does not mean the output is correct, and nothing here reproduces the paper's results. A line beginning “fixture:” under a sample that ran gives the arguments Syntology's fixture called the sample with, each as name:type/rank/element kind (? where none was recorded), and the type of what the fixture's run returned, exactly as the harness recorded them; the arguments and the returned type were recorded by two separate runs of the same fixture. No output value is shown, and a returned type says nothing about whether the output was right. “Level recorded” gives the UTC date when Syntology's graph wrote the sample's verification level, not when the code executed; where the level was written before Syntology's graph kept that date, the row says “level recorded, date not recorded”. For 1 sample that ran, the graph records a signature this page does not show: for a class the fixture built without calling, the harness called the object with inputs it made up itself, so the recorded signature is the harness's probe, not the fixture's arguments. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself. “Our draft was wrong” and “the contract check could not run” are failures of Syntology's instrument, not of the code: the sample ran on Syntology's synthesized input, and then the drafted contract did not describe it, or the check itself raised an error (RAISES), lacked a dependency (DEP_MISSING) or ran out of time (TIMEOUT); hover the label for which. A metamorphic tier, where shown, is evidence of the output's shape, not that the sample implements the method.

Each sample ends with its code_sha256. A sample's code_sha256 is Syntology's identity for that exact code. It is printed in full (16 hexadecimal characters for every sample in this build), and get_code(code_sha256="…") takes it exactly as printed. An agent calls it on Syntology's MCP service (how to connect); click an identity to copy that call.

Repository labels, per sample. official repository: The archive marks this repository official for the paper. named in the paper: The archive records that the paper mentions this repository; it is not marked official. community (archive-listed): In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper. found in paper text by Syntology: Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted. community: Not in the archive's code links for this paper; a community repository Syntology harvested. Samples are grouped by repository: a repository marked official first, then by how directly the repository is tied to the paper, then in the archive's order; never by how many ran. Licence labels name the repository's licence as recorded at harvest. “Pointer only” means Syntology does not serve that copy's text, for one of six reasons: no licence file was found; the licence was not identified; the licence is recorded as permissive but that copy's record is not marked cleared; the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar); this paper's copy was not located, so its licence is not recorded; or this paper's copy is cleared, but the stored copy of the same code that Syntology serves is not. Some licences outside that list permit redistribution, such as WTFPL, and GPL-3.0 under its conditions; they are simply not on the list. Hover a licence label for the reason. File links open the file on GitHub at the default branch, which may have changed since the harvest.

A function or class name links to its page when the name occurs in the harvested code of at least 5 papers; the name is a string, not an identity. A census tag appears when Syntology's census ran implementations with that name on one shared input per bucket: it gives how many were compared, from how many papers, and how many distinct outputs they produced, and names this sample's bucket when this sample was one of them.

pauldaoudi/food community (archive-listed) 11 samples, 9 ran (4 instrument-failed) · 2 unverified
Discriminator in 42 papers agents/algos_utils/darc_utils/networks_manager.py ran fingerprinted MIT (permissive) · 85dcfc68165685af · report fixture: returned Discriminator · level recorded 2026-09-22
ExpertDataset agents/algos_utils/darc_utils/networks_manager.py ran MIT (permissive) · 3072f035b0c85c12 · report fixture: returned ExpertDataset · level recorded 2026-09-22
RolloutStorage agents/algos_utils/darc_utils/networks_manager.py ran MIT (permissive) · d5c0698f7ffe2457 · report fixture: returned RolloutStorage · level recorded 2026-09-22
RunningMeanStd in 13 papers agents/algos_utils/darc_utils/networks_manager.py ran MIT (permissive) · 54a2a67d3023c95b · report fixture: returned RunningMeanStd · level recorded 2026-09-22
_flatten_helper agents/algos_utils/darc_utils/networks_manager.py ran · our draft was wrong fingerprinted MIT (permissive) · 0207b2d8ce743999 · report fixture: signature T:int/0/int, N:int/0/int, _tensor:Tensor/4/float · returned Tensor · level recorded 2026-09-22
expert_data_loader agents/algos_utils/darc_utils/networks_manager.py ran · our draft was wrong MIT (permissive) · 6e7428e76d042cba · report fixture: signature dataset:ExpertDataset/0/?, drop_last:bool/0/bool, num_grad_steps:int/0/int · returned DataLoader · level recorded 2026-09-22
get_dataset in 285 papers agents/algos_utils/darc_utils/networks_manager.py ran · our draft was wrong MIT (permissive) · c49250e1563ad2cd · report fixture: signature filename:str/0/str, num_trajs:int/0/int, subsample_frequency:int/0/int, to_add:str/0/str · returned ExpertDataset · level recorded 2026-09-22
obs_batch_normalize agents/algos_utils/darc_utils/networks_manager.py ran · our draft was wrong MIT (permissive) · 0642a6a4bea7b65a · report fixture: signature obs_tnsr:Tensor/2/float, update_rms:bool/0/bool, rms_obj:RunningMeanStd/0/? · returned Tensor · level recorded 2026-09-22
update_mean_var_count_from_moments in 14 papers agents/algos_utils/darc_utils/networks_manager.py ran · honoured contract fingerprinted MIT (permissive) · e2024afc1a6905ff · report fixture: signature mean:ndarray/1/float, var:ndarray/1/float, count:float/0/float, batch_mean:ndarray/1/float, batch_var:ndarray/1/float, batch_count:int/0/int · returned tuple · level recorded 2026-09-22
ConcatDiscriminator agents/algos_utils/darc_utils/networks_manager.py unverified MIT (permissive) · 4bf00ffb22a14dca · report
NetworksManager agents/algos_utils/darc_utils/networks_manager.py unverified MIT (permissive) · a1e7649f2401900f · report
shreyasc-13/off_domain_rlkit community (archive-listed) 6 samples, 6 ran (1 instrument-failed) · 0 unverified
clip_magnitude in 16 papers rlkit/pythonplusplus.py ran · honoured contract fingerprinted MIT (permissive) here; the stored copy is pointer only · 93d28ca67303e04b · report fixture: signature value:float/0/float, magnitude:float/0/float · returned float · level recorded 2026-09-24
create_stats_ordered_dict in 25 papers rlkit/core/eval_util.py ran MIT (permissive) here; the stored copy is pointer only · fadf5f762ff04cd6 · report fixture: signature name:str/0/str, data:ndarray/1/float, stat_prefix:str/0/str, always_show_all_stats:bool/0/bool, exclude_max_min:bool/0/bool · returned OrderedDict · level recorded 2026-09-24
get_average_returns in 27 papers rlkit/core/eval_util.py ran · honoured contract MIT (permissive) here; the stored copy is pointer only · a5d10a0b704c8db7 · report fixture: signature paths:list/1/dict · returned float64 · level recorded 2026-09-24
identity in 174 papers
census: 11 implementations from 39 papers, 3 distinct outputs across 2 buckets, one shared input per bucket this sample: bucket 1 of 2, arg 1: rank 2, kind float, dtype float32
rlkit/pythonplusplus.py ran · honoured contract fingerprinted MIT (permissive) here; the stored copy is pointer only · 9910e2fc297f8665 · report fixture: signature x:Tensor/2/float · returned Tensor · level recorded 2026-09-22
is_numeric in 28 papers rlkit/pythonplusplus.py ran · violated contract fingerprinted MIT (permissive) here; the stored copy is pointer only · 4138766e854db009 · report fixture: signature x:int/0/int · returned bool · level recorded 2026-09-24
simple_separated_format in 28 papers rlkit/core/tabulate.py ran · our draft was wrong MIT (permissive) here; the stored copy is pointer only · 5cb0c55f58279f59 · report fixture: signature separator:str/0/str · returned TableFormat · level recorded 2026-09-24
kchua/handful-of-trials found in paper text by Syntology 2 samples, 2 ran (0 instrument-failed) · 0 unverified
create_read_only dmbrl/config/default.py ran MIT (permissive) · 738c96c362e9ea1d · report fixture: signature message:str/0/str · returned function · level recorded 2026-09-25
make_bool dmbrl/config/default.py ran MIT (permissive) · 22eb5e3f463c48d7 · report fixture: signature arg:str/0/str · returned bool · level recorded 2026-09-25
JannerM/mbpo found in paper text by Syntology 4 samples, 4 ran (2 instrument-failed) · 0 unverified
td_target in 11 papers
census: 2 implementations from 6 papers, 2 distinct outputs across 2 buckets, one shared input per bucket this sample: bucket 2 of 2, arg 1: rank 2, kind float, dtype float32 · arg 2: rank 2, kind float, dtype float32
mbpo/algorithms/mbpo.py ran · honoured contract fingerprinted MIT (permissive) · 31deb9c7be58e00c · report fixture: signature reward:Tensor/2/float, discount:float/0/float, next_value:Tensor/2/float · returned Tensor · level recorded, date not recorded
format_samples_for_training in 7 papers mbpo/models/constructor.py ran · our draft was wrong MIT (permissive) · 1bb5583dba32f555 · report fixture: signature samples:dict/0/? · returned tuple · level recorded 2026-09-22
get_required_argument in 7 papers mbpo/models/utils.py ran · our draft was wrong MIT (permissive) here; the stored copy is pointer only · 4eb5226056a1f70f · report fixture: signature dotmap:dict/0/?, key:str/0/str, message:str/0/str, default:NoneType/0/? · returned str · level recorded 2026-09-22
mass_center in 8 papers mbpo/env/humanoid.py ran MIT (permissive) · 46bd205700c42f20 · report fixture: signature model:MockModel/0/?, sim:MockSim/0/? · returned float64 · level recorded 2026-09-25

Method as code Syntology

What Syntology's graph records as this paper's claimed method, lifted into code: each assembly names its bricks (functions and classes), each with its own role and its own verification level. An assembly is not a check that the code implements the method correctly or reproduces the paper's results. Levels are each brick's own verification level in Syntology's graph (level 0: no recorded run), listed brick by brick and never combined into one level; “ran” means the brick executed on a synthesized input, not that its output is correct, and the instrument figure counts failures of Syntology's instrument, not of the code. For a generated assembly, an independent peer is a second implementation generated separately from the same paper text; neither implementation read the authors' code. At level 2 the two agreed numerically on the fixture's inputs; level 3 is recorded when property tests derived from the paper also passed on both, and each level 3 card prints how many its report lists. Where a model read the code against the paper's claim, each assembly's label verdict is one model's judgement, kept in this page's JSON twin; it is not calibrated, so no card prints it. A generated assembly's label was set when its code was generated for the method, and no model judged it; its card prints it as "by construction" or "not judged by a model". Where an assembly's phrase has no constructed figure, a brick that ran is not among the harvested samples this build reads, and whether a brick's run only constructed an object is read from that brick's harvested sample, so the figure was not counted; it does not say there were none.

2 assemblies on this paper: 1 harvested, 1 generated.

generated Assembly for Off-Dynamics Reinforcement Learning (graph method: Off-Dynamics Reinforcement Learning), generated from the paper text, not the repository's code. 1 brick; individually on a fixture: 1 ran (1 with no instrument failure: 0 honoured, 1 violated, 0 with no contract checked; 0 where Syntology's instrument failed) · 0 unverified. Ran end-to-end against an independent peer (level 3; 4 property tests passed on both).

label matches by construction: the code was generated for this method (not judged by a model)

  1. compute_darc_reward_bonus entry · level 3 · ran · violated contract · 3d55c473ad8bdd85204f7c984f30049fa8203d3dbed7ea12fcae37aabbb791d5

harvested Assembly for Off-Dynamics Reinforcement Learning (graph method: Off-Dynamics Reinforcement Learning). 11 bricks; individually on a fixture: 9 ran (of which 4 constructed an object rather than computing a result; 5 with no instrument failure: 1 honoured, 0 violated, 4 with no contract checked; 4 where Syntology's instrument failed) · 2 unverified. The assembly has not been run end-to-end.

  1. NetworksManager entry · level 0 · unverified · a1e7649f2401900f
  2. ConcatDiscriminator model · level 0 · unverified · 4bf00ffb22a14dca
  3. RolloutStorage helper · level 1 · ran · d5c0698f7ffe2457
  4. RunningMeanStd helper · level 1 · ran · 54a2a67d3023c95b
  5. expert_data_loader helper · level 1 · ran · our draft was wrong · 6e7428e76d042cba
  6. get_dataset helper · level 1 · ran · our draft was wrong · c49250e1563ad2cd
  7. obs_batch_normalize helper · level 1 · ran · our draft was wrong · 0642a6a4bea7b65a
  8. Discriminator model · level 1 · ran · 85dcfc68165685af
  9. _flatten_helper helper · level 1 · ran · our draft was wrong · 0207b2d8ce743999
  10. update_mean_var_count_from_moments helper · level 1 · ran · honoured contract · e2024afc1a6905ff
  11. ExpertDataset helper · level 1 · ran · 3072f035b0c85c12

Each brick ends with its code_sha256, Syntology's identity for that exact code; click one to copy the get_code call that returns it. For agents, on Syntology's MCP service: assembly(arxiv_id="2006.13916") returns these assemblies with each brick's role and wiring (how to connect).

Tasks

Continuous ControlDomain AdaptationReinforcement LearningReinforcement Learning (RL)continuous-controlreinforcement-learning

Results from the paper archive 2025-07-28

No leaderboard rows for this paper in the archive.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections