{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2020:4NDRFIGWIJGZ5KSDCALEADONMT","short_pith_number":"pith:4NDRFIGW","schema_version":"1.0","canonical_sha256":"e34712a0d6424d9eaa431016400dcd64c2b37b157942c8ae6409a30949e3d0e7","source":{"kind":"arxiv","id":"2006.04678","version":2},"attestation_state":"computed","paper":{"title":"Primal Wasserstein Imitation Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"L\\'eonard Hussenot, Matthieu Geist, Olivier Pietquin, Robert Dadashi","submitted_at":"2020-06-08T15:30:11Z","abstract_excerpt":"Imitation Learning (IL) methods seek to match the behavior of an agent with that of an expert. In the present work, we propose a new IL method based on a conceptually simple algorithm: Primal Wasserstein Imitation Learning (PWIL), which ties to the primal form of the Wasserstein distance between the expert and the agent state-action distributions. We present a reward function which is derived offline, as opposed to recent adversarial IL algorithms that learn a reward function through interactions with the environment, and which requires little fine-tuning. We show that we can recover expert be"},"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":"2006.04678","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2020-06-08T15:30:11Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"e4ad8cba824c10b43d7606266d79d6e25f67c3113e9510a562eeca400272527d","abstract_canon_sha256":"4bfe6e1b1eef6cfd58fec6f48cc361d54038b0be26e720a75acdfe98e231d4d3"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:23:57.229851Z","signature_b64":"oQGo33UPKTD8I85BYs9L0MyVue2q4gdZD/hFXFD2ch8RexmIGnRjwQO1d8Q9mt3tpf9WeA8ZOlcqGzoY1GAJCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e34712a0d6424d9eaa431016400dcd64c2b37b157942c8ae6409a30949e3d0e7","last_reissued_at":"2026-07-05T02:23:57.229385Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:23:57.229385Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Primal Wasserstein Imitation Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"L\\'eonard Hussenot, Matthieu Geist, Olivier Pietquin, Robert Dadashi","submitted_at":"2020-06-08T15:30:11Z","abstract_excerpt":"Imitation Learning (IL) methods seek to match the behavior of an agent with that of an expert. In the present work, we propose a new IL method based on a conceptually simple algorithm: Primal Wasserstein Imitation Learning (PWIL), which ties to the primal form of the Wasserstein distance between the expert and the agent state-action distributions. We present a reward function which is derived offline, as opposed to recent adversarial IL algorithms that learn a reward function through interactions with the environment, and which requires little fine-tuning. We show that we can recover expert be"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2006.04678","kind":"arxiv","version":2},"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/2006.04678/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":"2006.04678","created_at":"2026-07-05T02:23:57.229446+00:00"},{"alias_kind":"arxiv_version","alias_value":"2006.04678v2","created_at":"2026-07-05T02:23:57.229446+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2006.04678","created_at":"2026-07-05T02:23:57.229446+00:00"},{"alias_kind":"pith_short_12","alias_value":"4NDRFIGWIJGZ","created_at":"2026-07-05T02:23:57.229446+00:00"},{"alias_kind":"pith_short_16","alias_value":"4NDRFIGWIJGZ5KSD","created_at":"2026-07-05T02:23:57.229446+00:00"},{"alias_kind":"pith_short_8","alias_value":"4NDRFIGW","created_at":"2026-07-05T02:23:57.229446+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.09381","citing_title":"ReGIL: Retrieval-Guided Imitation Learning from a Single Demonstration","ref_index":25,"is_internal_anchor":false},{"citing_arxiv_id":"2509.26627","citing_title":"TimeRewarder: Learning Dense Reward from Passive Videos via Frame-wise Temporal Distance","ref_index":3,"is_internal_anchor":false},{"citing_arxiv_id":"2604.13788","citing_title":"Failure Identification in Imitation Learning Via Statistical and Semantic Filtering","ref_index":51,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/4NDRFIGWIJGZ5KSDCALEADONMT","json":"https://pith.science/pith/4NDRFIGWIJGZ5KSDCALEADONMT.json","graph_json":"https://pith.science/api/pith-number/4NDRFIGWIJGZ5KSDCALEADONMT/graph.json","events_json":"https://pith.science/api/pith-number/4NDRFIGWIJGZ5KSDCALEADONMT/events.json","paper":"https://pith.science/paper/4NDRFIGW"},"agent_actions":{"view_html":"https://pith.science/pith/4NDRFIGWIJGZ5KSDCALEADONMT","download_json":"https://pith.science/pith/4NDRFIGWIJGZ5KSDCALEADONMT.json","view_paper":"https://pith.science/paper/4NDRFIGW","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2006.04678&json=true","fetch_graph":"https://pith.science/api/pith-number/4NDRFIGWIJGZ5KSDCALEADONMT/graph.json","fetch_events":"https://pith.science/api/pith-number/4NDRFIGWIJGZ5KSDCALEADONMT/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/4NDRFIGWIJGZ5KSDCALEADONMT/action/timestamp_anchor","attest_storage":"https://pith.science/pith/4NDRFIGWIJGZ5KSDCALEADONMT/action/storage_attestation","attest_author":"https://pith.science/pith/4NDRFIGWIJGZ5KSDCALEADONMT/action/author_attestation","sign_citation":"https://pith.science/pith/4NDRFIGWIJGZ5KSDCALEADONMT/action/citation_signature","submit_replication":"https://pith.science/pith/4NDRFIGWIJGZ5KSDCALEADONMT/action/replication_record"}},"created_at":"2026-07-05T02:23:57.229446+00:00","updated_at":"2026-07-05T02:23:57.229446+00:00"}