{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:5IH6CBLM2ZL7VNJ3I5E56YK2R3","short_pith_number":"pith:5IH6CBLM","schema_version":"1.0","canonical_sha256":"ea0fe1056cd657fab53b4749df615a8eead61f27b5f46b58855a78a92776573b","source":{"kind":"arxiv","id":"2507.23698","version":1},"attestation_state":"computed","paper":{"title":"Scalable Multi-Task Reinforcement Learning for Generalizable Spatial Intelligence in Visuomotor Agents","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.RO","authors_text":"Anji Liu, Bowei Zhang, Haiwen Xia, Shaofei Cai, Yitao Liang, Zhancun Mu","submitted_at":"2025-07-31T16:20:02Z","abstract_excerpt":"While Reinforcement Learning (RL) has achieved remarkable success in language modeling, its triumph hasn't yet fully translated to visuomotor agents. A primary challenge in RL models is their tendency to overfit specific tasks or environments, thereby hindering the acquisition of generalizable behaviors across diverse settings. This paper provides a preliminary answer to this challenge by demonstrating that RL-finetuned visuomotor agents in Minecraft can achieve zero-shot generalization to unseen worlds. Specifically, we explore RL's potential to enhance generalizable spatial reasoning and 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":"2507.23698","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.RO","submitted_at":"2025-07-31T16:20:02Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"78056bd75a226e1068ff7f7c75cc97480d2870e48c2c966f5dba1cbecd43e5d1","abstract_canon_sha256":"df3ffdff7b24888714be3134009b6c6b6a3e9ae5ea23ff9efe0552bd9c6eb99a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:46:29.525732Z","signature_b64":"NKh5sx8T3rsZuie0zwnRJpQZ06Q86Rt8s+vifx/AkK7Qdngipjtwp9Ahke3Vb5RBxLTOYVInL1rSey5paMuYAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ea0fe1056cd657fab53b4749df615a8eead61f27b5f46b58855a78a92776573b","last_reissued_at":"2026-07-05T11:46:29.525169Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:46:29.525169Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Scalable Multi-Task Reinforcement Learning for Generalizable Spatial Intelligence in Visuomotor Agents","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.RO","authors_text":"Anji Liu, Bowei Zhang, Haiwen Xia, Shaofei Cai, Yitao Liang, Zhancun Mu","submitted_at":"2025-07-31T16:20:02Z","abstract_excerpt":"While Reinforcement Learning (RL) has achieved remarkable success in language modeling, its triumph hasn't yet fully translated to visuomotor agents. A primary challenge in RL models is their tendency to overfit specific tasks or environments, thereby hindering the acquisition of generalizable behaviors across diverse settings. This paper provides a preliminary answer to this challenge by demonstrating that RL-finetuned visuomotor agents in Minecraft can achieve zero-shot generalization to unseen worlds. Specifically, we explore RL's potential to enhance generalizable spatial reasoning and int"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.23698","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/2507.23698/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":"2507.23698","created_at":"2026-07-05T11:46:29.525226+00:00"},{"alias_kind":"arxiv_version","alias_value":"2507.23698v1","created_at":"2026-07-05T11:46:29.525226+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.23698","created_at":"2026-07-05T11:46:29.525226+00:00"},{"alias_kind":"pith_short_12","alias_value":"5IH6CBLM2ZL7","created_at":"2026-07-05T11:46:29.525226+00:00"},{"alias_kind":"pith_short_16","alias_value":"5IH6CBLM2ZL7VNJ3","created_at":"2026-07-05T11:46:29.525226+00:00"},{"alias_kind":"pith_short_8","alias_value":"5IH6CBLM","created_at":"2026-07-05T11:46:29.525226+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.20246","citing_title":"GROW: Aligning GRPO with State-Action Modeling for Open-World VLM Agents","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2605.20246","citing_title":"GROW: Aligning GRPO with State-Action Modeling for Open-World VLM Agents","ref_index":5,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/5IH6CBLM2ZL7VNJ3I5E56YK2R3","json":"https://pith.science/pith/5IH6CBLM2ZL7VNJ3I5E56YK2R3.json","graph_json":"https://pith.science/api/pith-number/5IH6CBLM2ZL7VNJ3I5E56YK2R3/graph.json","events_json":"https://pith.science/api/pith-number/5IH6CBLM2ZL7VNJ3I5E56YK2R3/events.json","paper":"https://pith.science/paper/5IH6CBLM"},"agent_actions":{"view_html":"https://pith.science/pith/5IH6CBLM2ZL7VNJ3I5E56YK2R3","download_json":"https://pith.science/pith/5IH6CBLM2ZL7VNJ3I5E56YK2R3.json","view_paper":"https://pith.science/paper/5IH6CBLM","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2507.23698&json=true","fetch_graph":"https://pith.science/api/pith-number/5IH6CBLM2ZL7VNJ3I5E56YK2R3/graph.json","fetch_events":"https://pith.science/api/pith-number/5IH6CBLM2ZL7VNJ3I5E56YK2R3/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/5IH6CBLM2ZL7VNJ3I5E56YK2R3/action/timestamp_anchor","attest_storage":"https://pith.science/pith/5IH6CBLM2ZL7VNJ3I5E56YK2R3/action/storage_attestation","attest_author":"https://pith.science/pith/5IH6CBLM2ZL7VNJ3I5E56YK2R3/action/author_attestation","sign_citation":"https://pith.science/pith/5IH6CBLM2ZL7VNJ3I5E56YK2R3/action/citation_signature","submit_replication":"https://pith.science/pith/5IH6CBLM2ZL7VNJ3I5E56YK2R3/action/replication_record"}},"created_at":"2026-07-05T11:46:29.525226+00:00","updated_at":"2026-07-05T11:46:29.525226+00:00"}