{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:AGSOC5B6KJADFU455HTE4BVM64","short_pith_number":"pith:AGSOC5B6","schema_version":"1.0","canonical_sha256":"01a4e1743e524032d39de9e64e06acf728d74df50aeff6d62bf1ca706f6b1fe4","source":{"kind":"arxiv","id":"2502.04037","version":2},"attestation_state":"computed","paper":{"title":"Exploring Imbalanced Annotations for Effective In-Context Learning","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Bingyi Jing, Deyu Meng, Feipeng Zhang, Hao Zeng, Hongfu Gao, Hongxin Wei","submitted_at":"2025-02-06T12:57:50Z","abstract_excerpt":"Large language models (LLMs) have shown impressive performance on downstream tasks through in-context learning (ICL), which heavily relies on the demonstrations selected from annotated datasets. However, these datasets often exhibit long-tailed class distributions in real-world scenarios, leading to biased demonstration selection. In this work, we show that such class imbalances significantly degrade the ICL performance across various tasks, regardless of selection methods. Moreover, classical rebalancing methods, which focus solely on class weights, yield poor performance due to neglecting co"},"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":"2502.04037","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2025-02-06T12:57:50Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"925d22a43e6c33691edfe948d360ae5a598066e0186df345e28c1a984bab64bd","abstract_canon_sha256":"072f4b05c28b3795d1092eb3e692c95e2627de635395b4edfe41caf891845be9"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:12:25.553689Z","signature_b64":"DvMfwLDsgX73NLcsZfSW6ZeXBGqTlpknB5gA44qh1II7i87qz1nAbjR9cJD9uj6LJWTe2+MmJutUfBMO+SR1BA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"01a4e1743e524032d39de9e64e06acf728d74df50aeff6d62bf1ca706f6b1fe4","last_reissued_at":"2026-07-05T11:12:25.553060Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:12:25.553060Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Exploring Imbalanced Annotations for Effective In-Context Learning","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Bingyi Jing, Deyu Meng, Feipeng Zhang, Hao Zeng, Hongfu Gao, Hongxin Wei","submitted_at":"2025-02-06T12:57:50Z","abstract_excerpt":"Large language models (LLMs) have shown impressive performance on downstream tasks through in-context learning (ICL), which heavily relies on the demonstrations selected from annotated datasets. However, these datasets often exhibit long-tailed class distributions in real-world scenarios, leading to biased demonstration selection. In this work, we show that such class imbalances significantly degrade the ICL performance across various tasks, regardless of selection methods. Moreover, classical rebalancing methods, which focus solely on class weights, yield poor performance due to neglecting co"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.04037","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/2502.04037/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":"2502.04037","created_at":"2026-07-05T11:12:25.553143+00:00"},{"alias_kind":"arxiv_version","alias_value":"2502.04037v2","created_at":"2026-07-05T11:12:25.553143+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.04037","created_at":"2026-07-05T11:12:25.553143+00:00"},{"alias_kind":"pith_short_12","alias_value":"AGSOC5B6KJAD","created_at":"2026-07-05T11:12:25.553143+00:00"},{"alias_kind":"pith_short_16","alias_value":"AGSOC5B6KJADFU45","created_at":"2026-07-05T11:12:25.553143+00:00"},{"alias_kind":"pith_short_8","alias_value":"AGSOC5B6","created_at":"2026-07-05T11:12:25.553143+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/AGSOC5B6KJADFU455HTE4BVM64","json":"https://pith.science/pith/AGSOC5B6KJADFU455HTE4BVM64.json","graph_json":"https://pith.science/api/pith-number/AGSOC5B6KJADFU455HTE4BVM64/graph.json","events_json":"https://pith.science/api/pith-number/AGSOC5B6KJADFU455HTE4BVM64/events.json","paper":"https://pith.science/paper/AGSOC5B6"},"agent_actions":{"view_html":"https://pith.science/pith/AGSOC5B6KJADFU455HTE4BVM64","download_json":"https://pith.science/pith/AGSOC5B6KJADFU455HTE4BVM64.json","view_paper":"https://pith.science/paper/AGSOC5B6","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2502.04037&json=true","fetch_graph":"https://pith.science/api/pith-number/AGSOC5B6KJADFU455HTE4BVM64/graph.json","fetch_events":"https://pith.science/api/pith-number/AGSOC5B6KJADFU455HTE4BVM64/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/AGSOC5B6KJADFU455HTE4BVM64/action/timestamp_anchor","attest_storage":"https://pith.science/pith/AGSOC5B6KJADFU455HTE4BVM64/action/storage_attestation","attest_author":"https://pith.science/pith/AGSOC5B6KJADFU455HTE4BVM64/action/author_attestation","sign_citation":"https://pith.science/pith/AGSOC5B6KJADFU455HTE4BVM64/action/citation_signature","submit_replication":"https://pith.science/pith/AGSOC5B6KJADFU455HTE4BVM64/action/replication_record"}},"created_at":"2026-07-05T11:12:25.553143+00:00","updated_at":"2026-07-05T11:12:25.553143+00:00"}