{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2018:PL2DBDWXB3MREKKRGUVC3BDFIA","short_pith_number":"pith:PL2DBDWX","schema_version":"1.0","canonical_sha256":"7af4308ed70ed9122951352a2d8465401ebe3df61e3cd3f9a7626005c6b84c6f","source":{"kind":"arxiv","id":"1812.00420","version":2},"attestation_state":"computed","paper":{"title":"Efficient Lifelong Learning with A-GEM","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Arslan Chaudhry, Marc'Aurelio Ranzato, Marcus Rohrbach, Mohamed Elhoseiny","submitted_at":"2018-12-02T16:39:19Z","abstract_excerpt":"In lifelong learning, the learner is presented with a sequence of tasks, incrementally building a data-driven prior which may be leveraged to speed up learning of a new task. In this work, we investigate the efficiency of current lifelong approaches, in terms of sample complexity, computational and memory cost. Towards this end, we first introduce a new and a more realistic evaluation protocol, whereby learners observe each example only once and hyper-parameter selection is done on a small and disjoint set of tasks, which is not used for the actual learning experience and evaluation. Second, w"},"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":"1812.00420","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2018-12-02T16:39:19Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"6b8197e607c7365928c44bc359a33b5f55b0e66949d9c160c6b2155fdf0e311b","abstract_canon_sha256":"b43b2ae992172559bbef6105d5cf418cf2f09eb428a45d308754c7cc5f996b3e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:56:40.742342Z","signature_b64":"s17bUkdzY+YDT4PQemf+WvqKDd0ngeOjvl4cVeijtW5bbCv7F48rpnWc0I73lC7pa9cE4ljasUUBcKSN5c4MDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7af4308ed70ed9122951352a2d8465401ebe3df61e3cd3f9a7626005c6b84c6f","last_reissued_at":"2026-05-17T23:56:40.741713Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:56:40.741713Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Efficient Lifelong Learning with A-GEM","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Arslan Chaudhry, Marc'Aurelio Ranzato, Marcus Rohrbach, Mohamed Elhoseiny","submitted_at":"2018-12-02T16:39:19Z","abstract_excerpt":"In lifelong learning, the learner is presented with a sequence of tasks, incrementally building a data-driven prior which may be leveraged to speed up learning of a new task. In this work, we investigate the efficiency of current lifelong approaches, in terms of sample complexity, computational and memory cost. Towards this end, we first introduce a new and a more realistic evaluation protocol, whereby learners observe each example only once and hyper-parameter selection is done on a small and disjoint set of tasks, which is not used for the actual learning experience and evaluation. Second, w"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1812.00420","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":""},"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":"1812.00420","created_at":"2026-05-17T23:56:40.741809+00:00"},{"alias_kind":"arxiv_version","alias_value":"1812.00420v2","created_at":"2026-05-17T23:56:40.741809+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1812.00420","created_at":"2026-05-17T23:56:40.741809+00:00"},{"alias_kind":"pith_short_12","alias_value":"PL2DBDWXB3MR","created_at":"2026-05-18T12:32:46.962924+00:00"},{"alias_kind":"pith_short_16","alias_value":"PL2DBDWXB3MREKKR","created_at":"2026-05-18T12:32:46.962924+00:00"},{"alias_kind":"pith_short_8","alias_value":"PL2DBDWX","created_at":"2026-05-18T12:32:46.962924+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":18,"internal_anchor_count":9,"sample":[{"citing_arxiv_id":"2605.15384","citing_title":"Is One Score Enough? Rethinking the Evaluation of Sequentially Evolving LLM Memory","ref_index":3,"is_internal_anchor":true},{"citing_arxiv_id":"2605.08949","citing_title":"Muon-OGD: Muon-based Spectral Orthogonal Gradient Projection for LLM Continual Learning","ref_index":7,"is_internal_anchor":true},{"citing_arxiv_id":"2507.20997","citing_title":"Modular Delta Merging with Orthogonal Constraints: A Scalable Framework for Continual and Reversible Model Composition","ref_index":18,"is_internal_anchor":true},{"citing_arxiv_id":"2508.18187","citing_title":"BRAIN: Bias-Mitigation Continual Learning Approach to Vision-Brain Understanding","ref_index":69,"is_internal_anchor":true},{"citing_arxiv_id":"2511.01831","citing_title":"Routing-Based Continual Learning for Multimodal Large Language Models","ref_index":8,"is_internal_anchor":true},{"citing_arxiv_id":"2603.20410","citing_title":"SLE-FNO: Single-Layer Extensions for Task-Agnostic Continual Learning in Fourier Neural Operators","ref_index":63,"is_internal_anchor":true},{"citing_arxiv_id":"2605.12998","citing_title":"DRIFT: A Benchmark for Task-Free Continual Graph Learning with Continuous Distribution Shifts","ref_index":10,"is_internal_anchor":true},{"citing_arxiv_id":"2605.12998","citing_title":"DRIFT: A Benchmark for Task-Free Continual Graph Learning with Continuous Distribution Shifts","ref_index":10,"is_internal_anchor":true},{"citing_arxiv_id":"2605.13835","citing_title":"Unlocking Patch-Level Features for CLIP-Based Class-Incremental Learning","ref_index":5,"is_internal_anchor":true},{"citing_arxiv_id":"2605.11617","citing_title":"MIST: Reliable Streaming Decision Trees for Online Class-Incremental Learning via McDiarmid Bound","ref_index":13,"is_internal_anchor":false},{"citing_arxiv_id":"2306.03310","citing_title":"LIBERO: Benchmarking Knowledge Transfer for Lifelong Robot Learning","ref_index":12,"is_internal_anchor":false},{"citing_arxiv_id":"2605.08949","citing_title":"Muon-OGD: Muon-based Spectral Orthogonal Gradient Projection for LLM Continual Learning","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2605.05776","citing_title":"HEDP: A Hybrid Energy-Distance Prompt-based Framework for Domain Incremental Learning","ref_index":79,"is_internal_anchor":false},{"citing_arxiv_id":"2605.05732","citing_title":"CRAFT: Forgetting-Aware Intervention-Based Adaptation for Continual Learning","ref_index":2,"is_internal_anchor":false},{"citing_arxiv_id":"2605.05732","citing_title":"CRAFT: Forgetting-Aware Intervention-Based Adaptation for Continual Learning","ref_index":2,"is_internal_anchor":false},{"citing_arxiv_id":"2604.07266","citing_title":"Tracking Adaptation Time: Metrics for Temporal Distribution Shift","ref_index":11,"is_internal_anchor":false},{"citing_arxiv_id":"2604.15414","citing_title":"Beyond Single-Model Optimization: Preserving Plasticity in Continual Reinforcement Learning","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2605.02105","citing_title":"Sharpness-Aware Pretraining Mitigates Catastrophic Forgetting","ref_index":73,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/PL2DBDWXB3MREKKRGUVC3BDFIA","json":"https://pith.science/pith/PL2DBDWXB3MREKKRGUVC3BDFIA.json","graph_json":"https://pith.science/api/pith-number/PL2DBDWXB3MREKKRGUVC3BDFIA/graph.json","events_json":"https://pith.science/api/pith-number/PL2DBDWXB3MREKKRGUVC3BDFIA/events.json","paper":"https://pith.science/paper/PL2DBDWX"},"agent_actions":{"view_html":"https://pith.science/pith/PL2DBDWXB3MREKKRGUVC3BDFIA","download_json":"https://pith.science/pith/PL2DBDWXB3MREKKRGUVC3BDFIA.json","view_paper":"https://pith.science/paper/PL2DBDWX","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1812.00420&json=true","fetch_graph":"https://pith.science/api/pith-number/PL2DBDWXB3MREKKRGUVC3BDFIA/graph.json","fetch_events":"https://pith.science/api/pith-number/PL2DBDWXB3MREKKRGUVC3BDFIA/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/PL2DBDWXB3MREKKRGUVC3BDFIA/action/timestamp_anchor","attest_storage":"https://pith.science/pith/PL2DBDWXB3MREKKRGUVC3BDFIA/action/storage_attestation","attest_author":"https://pith.science/pith/PL2DBDWXB3MREKKRGUVC3BDFIA/action/author_attestation","sign_citation":"https://pith.science/pith/PL2DBDWXB3MREKKRGUVC3BDFIA/action/citation_signature","submit_replication":"https://pith.science/pith/PL2DBDWXB3MREKKRGUVC3BDFIA/action/replication_record"}},"created_at":"2026-05-17T23:56:40.741809+00:00","updated_at":"2026-05-17T23:56:40.741809+00:00"}