{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:D4SWBTLLJ3BQTBKPASMZHGOOKV","short_pith_number":"pith:D4SWBTLL","schema_version":"1.0","canonical_sha256":"1f2560cd6b4ec309854f04999399ce555f9710fae2bd0516b043108174a93c04","source":{"kind":"arxiv","id":"2505.11164","version":1},"attestation_state":"computed","paper":{"title":"Parkour in the Wild: Learning a General and Extensible Agile Locomotion Policy Using Multi-expert Distillation and RL Fine-tuning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.RO","authors_text":"Joshua Aurand, Junzhe He, Marco Hutter, Nikita Rudin","submitted_at":"2025-05-16T12:07:37Z","abstract_excerpt":"Legged robots are well-suited for navigating terrains inaccessible to wheeled robots, making them ideal for applications in search and rescue or space exploration. However, current control methods often struggle to generalize across diverse, unstructured environments. This paper introduces a novel framework for agile locomotion of legged robots by combining multi-expert distillation with reinforcement learning (RL) fine-tuning to achieve robust generalization. Initially, terrain-specific expert policies are trained to develop specialized locomotion skills. These policies are then distilled int"},"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":"2505.11164","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.RO","submitted_at":"2025-05-16T12:07:37Z","cross_cats_sorted":[],"title_canon_sha256":"b0d7a60e12c5d8127bc08c2e1fba8f9c951a54fce27a53d13dd6e025b1142043","abstract_canon_sha256":"c21b901422fe00ddbf9fde6bb8265553fef96b0740f7aae6abecf3a8774ffa95"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:04:12.028942Z","signature_b64":"+paPBYOF2AjYV8LIYm1vi0yLZs8XnddwJhSgg/xf8ENVc53gIJx00Uh0zaJgxTzKjen63VkZ2kTuNGXKnGHECQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"1f2560cd6b4ec309854f04999399ce555f9710fae2bd0516b043108174a93c04","last_reissued_at":"2026-07-05T11:04:12.028507Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:04:12.028507Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Parkour in the Wild: Learning a General and Extensible Agile Locomotion Policy Using Multi-expert Distillation and RL Fine-tuning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.RO","authors_text":"Joshua Aurand, Junzhe He, Marco Hutter, Nikita Rudin","submitted_at":"2025-05-16T12:07:37Z","abstract_excerpt":"Legged robots are well-suited for navigating terrains inaccessible to wheeled robots, making them ideal for applications in search and rescue or space exploration. However, current control methods often struggle to generalize across diverse, unstructured environments. This paper introduces a novel framework for agile locomotion of legged robots by combining multi-expert distillation with reinforcement learning (RL) fine-tuning to achieve robust generalization. Initially, terrain-specific expert policies are trained to develop specialized locomotion skills. These policies are then distilled int"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.11164","kind":"arxiv","version":1},"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/2505.11164/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":"2505.11164","created_at":"2026-07-05T11:04:12.028568+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.11164v1","created_at":"2026-07-05T11:04:12.028568+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.11164","created_at":"2026-07-05T11:04:12.028568+00:00"},{"alias_kind":"pith_short_12","alias_value":"D4SWBTLLJ3BQ","created_at":"2026-07-05T11:04:12.028568+00:00"},{"alias_kind":"pith_short_16","alias_value":"D4SWBTLLJ3BQTBKP","created_at":"2026-07-05T11:04:12.028568+00:00"},{"alias_kind":"pith_short_8","alias_value":"D4SWBTLL","created_at":"2026-07-05T11:04:12.028568+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":15,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.25765","citing_title":"StairMaster: Learning to Conquer Risky Hollow Stairs for Agile Quadrupedal Robots","ref_index":10,"is_internal_anchor":false},{"citing_arxiv_id":"2606.19928","citing_title":"SWAP: Symmetric Equivariant World-Model for Agile Robot Parkour","ref_index":10,"is_internal_anchor":false},{"citing_arxiv_id":"2606.31912","citing_title":"Learning Locomotion on Discrete Terrain via Minimal Proximity Sensing","ref_index":10,"is_internal_anchor":false},{"citing_arxiv_id":"2606.17833","citing_title":"HumanoidArena: Benchmarking Egocentric Hierarchical Whole-body Learning","ref_index":31,"is_internal_anchor":false},{"citing_arxiv_id":"2606.05880","citing_title":"TAGA: Terrain-aware Active Gaze Learning for Generalizable Agile Humanoid Locomotion","ref_index":30,"is_internal_anchor":false},{"citing_arxiv_id":"2606.05873","citing_title":"LadderMan: Learning Humanoid Perceptive Ladder Climbing","ref_index":39,"is_internal_anchor":false},{"citing_arxiv_id":"2606.04718","citing_title":"CoRe-MoE: Contrastive Reweighted Mixture of Experts for Multi-Terrain Humanoid Locomotion with Gait Adaptation","ref_index":29,"is_internal_anchor":false},{"citing_arxiv_id":"2606.31912","citing_title":"Learning Locomotion on Discrete Terrain via Minimal Proximity Sensing","ref_index":10,"is_internal_anchor":false},{"citing_arxiv_id":"2605.26478","citing_title":"Efficient On-policy Visual-RL via Stochastic Decoupled Policy Gradient","ref_index":12,"is_internal_anchor":false},{"citing_arxiv_id":"2605.30770","citing_title":"SSR: Scaling Surefooted and Symmetric Humanoid Traversal to the Open World","ref_index":26,"is_internal_anchor":false},{"citing_arxiv_id":"2605.21429","citing_title":"roto 2.0: The Robot Tactile Olympiad","ref_index":1,"is_internal_anchor":false},{"citing_arxiv_id":"2602.15827","citing_title":"Perceptive Humanoid Parkour: Chaining Dynamic Human Skills via Motion Matching","ref_index":33,"is_internal_anchor":false},{"citing_arxiv_id":"2604.19344","citing_title":"Quadruped Parkour Learning: Sparsely Gated Mixture of Experts with Visual Input","ref_index":22,"is_internal_anchor":false},{"citing_arxiv_id":"2511.04831","citing_title":"Isaac Lab: A GPU-Accelerated Simulation Framework for Multi-Modal Robot Learning","ref_index":86,"is_internal_anchor":false},{"citing_arxiv_id":"2605.07988","citing_title":"Evaluation of an Actuated Spine in Agile Quadruped Locomotion","ref_index":11,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/D4SWBTLLJ3BQTBKPASMZHGOOKV","json":"https://pith.science/pith/D4SWBTLLJ3BQTBKPASMZHGOOKV.json","graph_json":"https://pith.science/api/pith-number/D4SWBTLLJ3BQTBKPASMZHGOOKV/graph.json","events_json":"https://pith.science/api/pith-number/D4SWBTLLJ3BQTBKPASMZHGOOKV/events.json","paper":"https://pith.science/paper/D4SWBTLL"},"agent_actions":{"view_html":"https://pith.science/pith/D4SWBTLLJ3BQTBKPASMZHGOOKV","download_json":"https://pith.science/pith/D4SWBTLLJ3BQTBKPASMZHGOOKV.json","view_paper":"https://pith.science/paper/D4SWBTLL","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.11164&json=true","fetch_graph":"https://pith.science/api/pith-number/D4SWBTLLJ3BQTBKPASMZHGOOKV/graph.json","fetch_events":"https://pith.science/api/pith-number/D4SWBTLLJ3BQTBKPASMZHGOOKV/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/D4SWBTLLJ3BQTBKPASMZHGOOKV/action/timestamp_anchor","attest_storage":"https://pith.science/pith/D4SWBTLLJ3BQTBKPASMZHGOOKV/action/storage_attestation","attest_author":"https://pith.science/pith/D4SWBTLLJ3BQTBKPASMZHGOOKV/action/author_attestation","sign_citation":"https://pith.science/pith/D4SWBTLLJ3BQTBKPASMZHGOOKV/action/citation_signature","submit_replication":"https://pith.science/pith/D4SWBTLLJ3BQTBKPASMZHGOOKV/action/replication_record"}},"created_at":"2026-07-05T11:04:12.028568+00:00","updated_at":"2026-07-05T11:04:12.028568+00:00"}