{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:MYM37LYDFPLETIVJYGNTY55LSR","short_pith_number":"pith:MYM37LYD","schema_version":"1.0","canonical_sha256":"6619bfaf032bd649a2a9c19b3c77ab9465e4542e1f4cdf1b8140890ab37a8c8e","source":{"kind":"arxiv","id":"2307.03864","version":4},"attestation_state":"computed","paper":{"title":"When Do Transformers Shine in RL? Decoupling Memory from Credit Assignment","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Benjamin Eysenbach, Michel Ma, Pierre-Luc Bacon, Tianwei Ni","submitted_at":"2023-07-07T23:34:12Z","abstract_excerpt":"Reinforcement learning (RL) algorithms face two distinct challenges: learning effective representations of past and present observations, and determining how actions influence future returns. Both challenges involve modeling long-term dependencies. The Transformer architecture has been very successful to solve problems that involve long-term dependencies, including in the RL domain. However, the underlying reason for the strong performance of Transformer-based RL methods remains unclear: is it because they learn effective memory, or because they perform effective credit assignment? After intro"},"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":"2307.03864","kind":"arxiv","version":4},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2023-07-07T23:34:12Z","cross_cats_sorted":[],"title_canon_sha256":"a378c8f67dc5a132536a693e181c3f0fe0654285f027778c06178d1313c93eb0","abstract_canon_sha256":"c0ca19213d3f27b5ce96f350773375bf9c99e1b20bf03bcd10a3e91c4b327162"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:08:42.558690Z","signature_b64":"owyge1M/ok3zYReGgN3DTck0MKNIAh7nMWSlu8qvynNGtdK9KX9kgl6xjD5YFjCKK2k6y0Jq6zlr4k57kwNgCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6619bfaf032bd649a2a9c19b3c77ab9465e4542e1f4cdf1b8140890ab37a8c8e","last_reissued_at":"2026-07-05T07:08:42.558196Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:08:42.558196Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"When Do Transformers Shine in RL? Decoupling Memory from Credit Assignment","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Benjamin Eysenbach, Michel Ma, Pierre-Luc Bacon, Tianwei Ni","submitted_at":"2023-07-07T23:34:12Z","abstract_excerpt":"Reinforcement learning (RL) algorithms face two distinct challenges: learning effective representations of past and present observations, and determining how actions influence future returns. Both challenges involve modeling long-term dependencies. The Transformer architecture has been very successful to solve problems that involve long-term dependencies, including in the RL domain. However, the underlying reason for the strong performance of Transformer-based RL methods remains unclear: is it because they learn effective memory, or because they perform effective credit assignment? After intro"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2307.03864","kind":"arxiv","version":4},"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/2307.03864/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":"2307.03864","created_at":"2026-07-05T07:08:42.558257+00:00"},{"alias_kind":"arxiv_version","alias_value":"2307.03864v4","created_at":"2026-07-05T07:08:42.558257+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2307.03864","created_at":"2026-07-05T07:08:42.558257+00:00"},{"alias_kind":"pith_short_12","alias_value":"MYM37LYDFPLE","created_at":"2026-07-05T07:08:42.558257+00:00"},{"alias_kind":"pith_short_16","alias_value":"MYM37LYDFPLETIVJ","created_at":"2026-07-05T07:08:42.558257+00:00"},{"alias_kind":"pith_short_8","alias_value":"MYM37LYD","created_at":"2026-07-05T07:08:42.558257+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2411.11188","citing_title":"AMAGO-2: Breaking the Multi-Task Barrier in Meta-Reinforcement Learning with Transformers","ref_index":56,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/MYM37LYDFPLETIVJYGNTY55LSR","json":"https://pith.science/pith/MYM37LYDFPLETIVJYGNTY55LSR.json","graph_json":"https://pith.science/api/pith-number/MYM37LYDFPLETIVJYGNTY55LSR/graph.json","events_json":"https://pith.science/api/pith-number/MYM37LYDFPLETIVJYGNTY55LSR/events.json","paper":"https://pith.science/paper/MYM37LYD"},"agent_actions":{"view_html":"https://pith.science/pith/MYM37LYDFPLETIVJYGNTY55LSR","download_json":"https://pith.science/pith/MYM37LYDFPLETIVJYGNTY55LSR.json","view_paper":"https://pith.science/paper/MYM37LYD","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2307.03864&json=true","fetch_graph":"https://pith.science/api/pith-number/MYM37LYDFPLETIVJYGNTY55LSR/graph.json","fetch_events":"https://pith.science/api/pith-number/MYM37LYDFPLETIVJYGNTY55LSR/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/MYM37LYDFPLETIVJYGNTY55LSR/action/timestamp_anchor","attest_storage":"https://pith.science/pith/MYM37LYDFPLETIVJYGNTY55LSR/action/storage_attestation","attest_author":"https://pith.science/pith/MYM37LYDFPLETIVJYGNTY55LSR/action/author_attestation","sign_citation":"https://pith.science/pith/MYM37LYDFPLETIVJYGNTY55LSR/action/citation_signature","submit_replication":"https://pith.science/pith/MYM37LYDFPLETIVJYGNTY55LSR/action/replication_record"}},"created_at":"2026-07-05T07:08:42.558257+00:00","updated_at":"2026-07-05T07:08:42.558257+00:00"}