{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/reward-machines-for-deep-rl-in-noisy-and","title":"Reward Machines for Deep RL in Noisy and Uncertain Environments","arxiv_id":"2406.00120","date":"2024-05-31","proceeding":null,"authors":["Andrew C. Li","Zizhao Chen","Toryn Q. Klassen","Pashootan Vaezipoor","Rodrigo Toro Icarte","Sheila A. McIlraith"],"abstract":"Reward Machines provide an automaton-inspired structure for specifying instructions, safety constraints, and other temporally extended reward-worthy behaviour. By exposing the underlying structure of a reward function, they enable the decomposition of an RL task, leading to impressive gains in sample efficiency. Although Reward Machines and similar formal specifications have a rich history of application towards sequential decision-making problems, they critically rely on a ground-truth interpretation of the domain-specific vocabulary that forms the building blocks of the reward function--such ground-truth interpretations are elusive in the real world due in part to partial observability and noisy sensing. In this work, we explore the use of Reward Machines for Deep RL in noisy and uncertain environments. We characterize this problem as a POMDP and propose a suite of RL algorithms that exploit task structure under uncertain interpretation of the domain-specific vocabulary. Through theory and experiments, we expose pitfalls in naive approaches to this problem while simultaneously demonstrating how task structure can be successfully leveraged under noisy interpretations of the vocabulary.","url_abs":"https://arxiv.org/abs/2406.00120v4","url_pdf":"https://arxiv.org/pdf/2406.00120v4.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"reward-machines-for-deep-rl-in-noisy-and","repo_url":"https://github.com/andrewli77/reward-machines-noisy-environments","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"decision-making","task_name":"Decision Making"},{"task_slug":"sequential-decision-making","task_name":"Sequential Decision Making"},{"task_slug":null,"task_name":"counterfactual"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/2406.00120","atlas_url":"https://app.syntology.ai/?focus=2406.00120","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2406.00120"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-25T09:33:49+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/andrewli77/reward-machines-noisy-environments","reach":{"status":"ok","spdx":"MIT"}}],"summary":{"ran":4,"unverified":2},"by_repo_kind":{"official":{"samples":6,"ran":4,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"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"},"n_pointer_only_for_licence":0,"samples":[{"code_sha256_prefix":"c7ccf6940026d778","entry":"evaluate_dnf","repo":"andrewli77/reward-machines-noisy-environments","repo_kind":"official","path":"src/reward_machines/reward_machine_utils.py","file_url":"https://github.com/andrewli77/reward-machines-noisy-environments/blob/HEAD/src/reward_machines/reward_machine_utils.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"c7ccf6940026d778"}},{"code_sha256_prefix":"08eabb75de3ea8cc","entry":"fc","repo":"andrewli77/reward-machines-noisy-environments","repo_kind":"official","path":"src/policy_network.py","file_url":"https://github.com/andrewli77/reward-machines-noisy-environments/blob/HEAD/src/policy_network.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"08eabb75de3ea8cc"}},{"code_sha256_prefix":"8ecbbfbc94998d2d","entry":"possible_true_propositions","repo":"andrewli77/reward-machines-noisy-environments","repo_kind":"official","path":"src/torch_ac/belief.py","file_url":"https://github.com/andrewli77/reward-machines-noisy-environments/blob/HEAD/src/torch_ac/belief.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"8ecbbfbc94998d2d"}},{"code_sha256_prefix":"f9ef1a9ba62a26bc","entry":"threshold_rm_beliefs","repo":"andrewli77/reward-machines-noisy-environments","repo_kind":"official","path":"src/torch_ac/belief.py","file_url":"https://github.com/andrewli77/reward-machines-noisy-environments/blob/HEAD/src/torch_ac/belief.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"f9ef1a9ba62a26bc"}},{"code_sha256_prefix":"5b5d3c0eb71a3aa2","entry":"getEnvModel","repo":"andrewli77/reward-machines-noisy-environments","repo_kind":"official","path":"src/env_model.py","file_url":"https://github.com/andrewli77/reward-machines-noisy-environments/blob/HEAD/src/env_model.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"5b5d3c0eb71a3aa2"}},{"code_sha256_prefix":"f1e0170748710dcd","entry":"value_iteration","repo":"andrewli77/reward-machines-noisy-environments","repo_kind":"official","path":"src/reward_machines/reward_machine_utils.py","file_url":"https://github.com/andrewli77/reward-machines-noisy-environments/blob/HEAD/src/reward_machines/reward_machine_utils.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"f1e0170748710dcd"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}