{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2023:KGXYUFTU5ALNC4EAQJIHCANMB2","short_pith_number":"pith:KGXYUFTU","canonical_record":{"source":{"id":"2306.04618","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2023-06-07T17:47:03Z","cross_cats_sorted":["cs.CR","cs.LG"],"title_canon_sha256":"a5c2361bbe76feca7f349ca70dc6d5e2016ea9b139f90d0b6dae66f815f5814f","abstract_canon_sha256":"1b12e3e83afb17c0d0a66c31dd3eaf55cd6c424664af548924dd0cefa7a1e60a"},"schema_version":"1.0"},"canonical_sha256":"51af8a1674e816d1708082507101ac0e9ebffc1dd1d4e62054abcaf6560fe886","source":{"kind":"arxiv","id":"2306.04618","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2306.04618","created_at":"2026-07-05T07:05:09Z"},{"alias_kind":"arxiv_version","alias_value":"2306.04618v2","created_at":"2026-07-05T07:05:09Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2306.04618","created_at":"2026-07-05T07:05:09Z"},{"alias_kind":"pith_short_12","alias_value":"KGXYUFTU5ALN","created_at":"2026-07-05T07:05:09Z"},{"alias_kind":"pith_short_16","alias_value":"KGXYUFTU5ALNC4EA","created_at":"2026-07-05T07:05:09Z"},{"alias_kind":"pith_short_8","alias_value":"KGXYUFTU","created_at":"2026-07-05T07:05:09Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2023:KGXYUFTU5ALNC4EAQJIHCANMB2","target":"record","payload":{"canonical_record":{"source":{"id":"2306.04618","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2023-06-07T17:47:03Z","cross_cats_sorted":["cs.CR","cs.LG"],"title_canon_sha256":"a5c2361bbe76feca7f349ca70dc6d5e2016ea9b139f90d0b6dae66f815f5814f","abstract_canon_sha256":"1b12e3e83afb17c0d0a66c31dd3eaf55cd6c424664af548924dd0cefa7a1e60a"},"schema_version":"1.0"},"canonical_sha256":"51af8a1674e816d1708082507101ac0e9ebffc1dd1d4e62054abcaf6560fe886","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:05:09.670499Z","signature_b64":"I21DkFhovIjtNs7PRXaqqc5MBpCnzn6EGqyr5skSIu21JoP9GvGID3gfp24hXUmCeuOnS6xzn0n2l+NoYI4ACg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"51af8a1674e816d1708082507101ac0e9ebffc1dd1d4e62054abcaf6560fe886","last_reissued_at":"2026-07-05T07:05:09.669992Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:05:09.669992Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2306.04618","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T07:05:09Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"5K/PymtuOyH4Qki27v8eoUREtd4XcxUaa8pDSjYH2cHHPhCNeHqqgoDOHRvp6A3RnhVDY35wYHlIaMbdiKE+Bg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-06T03:04:05.044272Z"},"content_sha256":"d3001f7e2236b581092d8f12372b96a66cec296eb447030e9232414db72ab4f7","schema_version":"1.0","event_id":"sha256:d3001f7e2236b581092d8f12372b96a66cec296eb447030e9232414db72ab4f7"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2023:KGXYUFTU5ALNC4EAQJIHCANMB2","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Revisiting Out-of-distribution Robustness in NLP: Benchmark, Analysis, and LLMs Evaluations","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.CR","cs.LG"],"primary_cat":"cs.CL","authors_text":"Fangyuan Zou, Ganqu Cui, Heng Ji, Hongcheng Gao, Lifan Yuan, Maosong Sun, Xingyi Cheng, Yangyi Chen, Zhiyuan Liu","submitted_at":"2023-06-07T17:47:03Z","abstract_excerpt":"This paper reexamines the research on out-of-distribution (OOD) robustness in the field of NLP. We find that the distribution shift settings in previous studies commonly lack adequate challenges, hindering the accurate evaluation of OOD robustness. To address these issues, we propose a benchmark construction protocol that ensures clear differentiation and challenging distribution shifts. Then we introduce BOSS, a Benchmark suite for Out-of-distribution robustneSS evaluation covering 5 tasks and 20 datasets. Based on BOSS, we conduct a series of experiments on pre-trained language models for an"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2306.04618","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/2306.04618/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T07:05:09Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"dI/vumdWUEKDM+RSWwfBuoig5Uzi0jEKDyM9VYCvClRAFVcDqp4T3Mss2uHH5XVJL7QHhYscmqt92AkSYbzDDA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-06T03:04:05.044899Z"},"content_sha256":"b478f67810bf28d9b40a420c18da553ea26b4795f872ad2c8e601ef231face18","schema_version":"1.0","event_id":"sha256:b478f67810bf28d9b40a420c18da553ea26b4795f872ad2c8e601ef231face18"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/KGXYUFTU5ALNC4EAQJIHCANMB2/bundle.json","state_url":"https://pith.science/pith/KGXYUFTU5ALNC4EAQJIHCANMB2/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/KGXYUFTU5ALNC4EAQJIHCANMB2/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-08-06T03:04:05Z","links":{"resolver":"https://pith.science/pith/KGXYUFTU5ALNC4EAQJIHCANMB2","bundle":"https://pith.science/pith/KGXYUFTU5ALNC4EAQJIHCANMB2/bundle.json","state":"https://pith.science/pith/KGXYUFTU5ALNC4EAQJIHCANMB2/state.json","well_known_bundle":"https://pith.science/.well-known/pith/KGXYUFTU5ALNC4EAQJIHCANMB2/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:KGXYUFTU5ALNC4EAQJIHCANMB2","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"1b12e3e83afb17c0d0a66c31dd3eaf55cd6c424664af548924dd0cefa7a1e60a","cross_cats_sorted":["cs.CR","cs.LG"],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2023-06-07T17:47:03Z","title_canon_sha256":"a5c2361bbe76feca7f349ca70dc6d5e2016ea9b139f90d0b6dae66f815f5814f"},"schema_version":"1.0","source":{"id":"2306.04618","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2306.04618","created_at":"2026-07-05T07:05:09Z"},{"alias_kind":"arxiv_version","alias_value":"2306.04618v2","created_at":"2026-07-05T07:05:09Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2306.04618","created_at":"2026-07-05T07:05:09Z"},{"alias_kind":"pith_short_12","alias_value":"KGXYUFTU5ALN","created_at":"2026-07-05T07:05:09Z"},{"alias_kind":"pith_short_16","alias_value":"KGXYUFTU5ALNC4EA","created_at":"2026-07-05T07:05:09Z"},{"alias_kind":"pith_short_8","alias_value":"KGXYUFTU","created_at":"2026-07-05T07:05:09Z"}],"graph_snapshots":[{"event_id":"sha256:b478f67810bf28d9b40a420c18da553ea26b4795f872ad2c8e601ef231face18","target":"graph","created_at":"2026-07-05T07:05:09Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2306.04618/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"This paper reexamines the research on out-of-distribution (OOD) robustness in the field of NLP. We find that the distribution shift settings in previous studies commonly lack adequate challenges, hindering the accurate evaluation of OOD robustness. To address these issues, we propose a benchmark construction protocol that ensures clear differentiation and challenging distribution shifts. Then we introduce BOSS, a Benchmark suite for Out-of-distribution robustneSS evaluation covering 5 tasks and 20 datasets. Based on BOSS, we conduct a series of experiments on pre-trained language models for an","authors_text":"Fangyuan Zou, Ganqu Cui, Heng Ji, Hongcheng Gao, Lifan Yuan, Maosong Sun, Xingyi Cheng, Yangyi Chen, Zhiyuan Liu","cross_cats":["cs.CR","cs.LG"],"headline":"","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2023-06-07T17:47:03Z","title":"Revisiting Out-of-distribution Robustness in NLP: Benchmark, Analysis, and LLMs Evaluations"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2306.04618","kind":"arxiv","version":2},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:d3001f7e2236b581092d8f12372b96a66cec296eb447030e9232414db72ab4f7","target":"record","created_at":"2026-07-05T07:05:09Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"1b12e3e83afb17c0d0a66c31dd3eaf55cd6c424664af548924dd0cefa7a1e60a","cross_cats_sorted":["cs.CR","cs.LG"],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2023-06-07T17:47:03Z","title_canon_sha256":"a5c2361bbe76feca7f349ca70dc6d5e2016ea9b139f90d0b6dae66f815f5814f"},"schema_version":"1.0","source":{"id":"2306.04618","kind":"arxiv","version":2}},"canonical_sha256":"51af8a1674e816d1708082507101ac0e9ebffc1dd1d4e62054abcaf6560fe886","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"51af8a1674e816d1708082507101ac0e9ebffc1dd1d4e62054abcaf6560fe886","first_computed_at":"2026-07-05T07:05:09.669992Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T07:05:09.669992Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"I21DkFhovIjtNs7PRXaqqc5MBpCnzn6EGqyr5skSIu21JoP9GvGID3gfp24hXUmCeuOnS6xzn0n2l+NoYI4ACg==","signature_status":"signed_v1","signed_at":"2026-07-05T07:05:09.670499Z","signed_message":"canonical_sha256_bytes"},"source_id":"2306.04618","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:d3001f7e2236b581092d8f12372b96a66cec296eb447030e9232414db72ab4f7","sha256:b478f67810bf28d9b40a420c18da553ea26b4795f872ad2c8e601ef231face18"],"state_sha256":"f3f4a7213aa4fed3205b7a2017a77a528dd4ada96fff60f559abbeb2bd8954e1"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"nMwrncfeIwE9BPLGngYv2tuu9uadh4n2jEFMkkQF/NP1iGfeIpWt/4fEyKxnbKmfZYpCnoNwhJTgZXedrz0KBA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-06T03:04:05.052435Z","bundle_sha256":"4ca5500629cd50cff298635c02d5ace0e5faf47572965d212f64b0c0273c4732"}}