{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:4JOQZFFKJXIMCB4F26JA2J3Q6E","short_pith_number":"pith:4JOQZFFK","schema_version":"1.0","canonical_sha256":"e25d0c94aa4dd0c10785d7920d2770f11dd89f9f608f94db7a8eebf5443d0de5","source":{"kind":"arxiv","id":"2312.07392","version":3},"attestation_state":"computed","paper":{"title":"ReRoGCRL: Representation-based Robustness in Goal-Conditioned Reinforcement Learning","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Jiaxu Liu, Meng Fang, Sihao Wu, Wenjie Ruan, Xiangyu Yin, Xiaowei Huang, Xingyu Zhao","submitted_at":"2023-12-12T16:05:55Z","abstract_excerpt":"While Goal-Conditioned Reinforcement Learning (GCRL) has gained attention, its algorithmic robustness against adversarial perturbations remains unexplored. The attacks and robust representation training methods that are designed for traditional RL become less effective when applied to GCRL. To address this challenge, we first propose the Semi-Contrastive Representation attack, a novel approach inspired by the adversarial contrastive attack. Unlike existing attacks in RL, it only necessitates information from the policy function and can be seamlessly implemented during deployment. Then, to miti"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2312.07392","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2023-12-12T16:05:55Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"15af61408baba860872cb09e600ad255c575c6f583bb2e01f6fdaabfbc3eeda3","abstract_canon_sha256":"466379f29a551e47357d56fb444168885fa70cec2302f8a4fb9830696071fe38"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:26:19.624240Z","signature_b64":"VHOtLZVmLYVpDG3f6u3mdTSZ+3Jgce8MhvyBKZ7IAAY6KcUrk0CksFJBUbdSWJY+MvFWKjbKe1gYdEKq4z9SCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e25d0c94aa4dd0c10785d7920d2770f11dd89f9f608f94db7a8eebf5443d0de5","last_reissued_at":"2026-07-05T07:26:19.623780Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:26:19.623780Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"ReRoGCRL: Representation-based Robustness in Goal-Conditioned Reinforcement Learning","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Jiaxu Liu, Meng Fang, Sihao Wu, Wenjie Ruan, Xiangyu Yin, Xiaowei Huang, Xingyu Zhao","submitted_at":"2023-12-12T16:05:55Z","abstract_excerpt":"While Goal-Conditioned Reinforcement Learning (GCRL) has gained attention, its algorithmic robustness against adversarial perturbations remains unexplored. The attacks and robust representation training methods that are designed for traditional RL become less effective when applied to GCRL. To address this challenge, we first propose the Semi-Contrastive Representation attack, a novel approach inspired by the adversarial contrastive attack. Unlike existing attacks in RL, it only necessitates information from the policy function and can be seamlessly implemented during deployment. Then, to miti"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2312.07392","kind":"arxiv","version":3},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2312.07392/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2312.07392","created_at":"2026-07-05T07:26:19.623838+00:00"},{"alias_kind":"arxiv_version","alias_value":"2312.07392v3","created_at":"2026-07-05T07:26:19.623838+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2312.07392","created_at":"2026-07-05T07:26:19.623838+00:00"},{"alias_kind":"pith_short_12","alias_value":"4JOQZFFKJXIM","created_at":"2026-07-05T07:26:19.623838+00:00"},{"alias_kind":"pith_short_16","alias_value":"4JOQZFFKJXIMCB4F","created_at":"2026-07-05T07:26:19.623838+00:00"},{"alias_kind":"pith_short_8","alias_value":"4JOQZFFK","created_at":"2026-07-05T07:26:19.623838+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.09364","citing_title":"Multi-scale Predictive Representations for Goal-conditioned Reinforcement Learning","ref_index":40,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/4JOQZFFKJXIMCB4F26JA2J3Q6E","json":"https://pith.science/pith/4JOQZFFKJXIMCB4F26JA2J3Q6E.json","graph_json":"https://pith.science/api/pith-number/4JOQZFFKJXIMCB4F26JA2J3Q6E/graph.json","events_json":"https://pith.science/api/pith-number/4JOQZFFKJXIMCB4F26JA2J3Q6E/events.json","paper":"https://pith.science/paper/4JOQZFFK"},"agent_actions":{"view_html":"https://pith.science/pith/4JOQZFFKJXIMCB4F26JA2J3Q6E","download_json":"https://pith.science/pith/4JOQZFFKJXIMCB4F26JA2J3Q6E.json","view_paper":"https://pith.science/paper/4JOQZFFK","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2312.07392&json=true","fetch_graph":"https://pith.science/api/pith-number/4JOQZFFKJXIMCB4F26JA2J3Q6E/graph.json","fetch_events":"https://pith.science/api/pith-number/4JOQZFFKJXIMCB4F26JA2J3Q6E/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/4JOQZFFKJXIMCB4F26JA2J3Q6E/action/timestamp_anchor","attest_storage":"https://pith.science/pith/4JOQZFFKJXIMCB4F26JA2J3Q6E/action/storage_attestation","attest_author":"https://pith.science/pith/4JOQZFFKJXIMCB4F26JA2J3Q6E/action/author_attestation","sign_citation":"https://pith.science/pith/4JOQZFFKJXIMCB4F26JA2J3Q6E/action/citation_signature","submit_replication":"https://pith.science/pith/4JOQZFFKJXIMCB4F26JA2J3Q6E/action/replication_record"}},"created_at":"2026-07-05T07:26:19.623838+00:00","updated_at":"2026-07-05T07:26:19.623838+00:00"}