{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2022:ACYTWDAEJ6VQHJIP7TOXNO7I7J","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":"ddd6936e5c6523ab1c622a20dfcf1e0538498f7a0ec3b71ac7e4be83ad23f93b","cross_cats_sorted":["cs.AI","cs.LG","cs.SI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2022-11-01T11:01:36Z","title_canon_sha256":"e751a54ed1818d160daa31bc65b714102bfef1d0fa479b91e7facd14dda7ece2"},"schema_version":"1.0","source":{"id":"2211.00384","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2211.00384","created_at":"2026-07-05T05:12:55Z"},{"alias_kind":"arxiv_version","alias_value":"2211.00384v2","created_at":"2026-07-05T05:12:55Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2211.00384","created_at":"2026-07-05T05:12:55Z"},{"alias_kind":"pith_short_12","alias_value":"ACYTWDAEJ6VQ","created_at":"2026-07-05T05:12:55Z"},{"alias_kind":"pith_short_16","alias_value":"ACYTWDAEJ6VQHJIP","created_at":"2026-07-05T05:12:55Z"},{"alias_kind":"pith_short_8","alias_value":"ACYTWDAE","created_at":"2026-07-05T05:12:55Z"}],"graph_snapshots":[{"event_id":"sha256:8ecf8ba3a8952fb211a4a8a641b2ff927035a5e9c8b0ab1877c97dfb7d486e08","target":"graph","created_at":"2026-07-05T05:12:55Z","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/2211.00384/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large pre-trained language models (LPLM) have shown spectacular success when fine-tuned on downstream supervised tasks. Yet, it is known that their performance can drastically drop when there is a distribution shift between the data used during training and that used at inference time. In this paper we focus on data distributions that naturally change over time and introduce four new REDDIT datasets, namely the WALLSTREETBETS, ASKSCIENCE, THE DONALD, and POLITICS sub-reddits. First, we empirically demonstrate that LPLM can display average performance drops of about 88% (in the best case!) when","authors_text":"C\\'esar Ojeda, Kostadin Cvejoski, Rams\\'es J. S\\'anchez","cross_cats":["cs.AI","cs.LG","cs.SI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2022-11-01T11:01:36Z","title":"The future is different: Large pre-trained language models fail in prediction tasks"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2211.00384","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":"sha256:94bf25adae03b592ed71e92a97f1b28db63c00c3fe913a778e6dbd4aa5603e29","target":"record","created_at":"2026-07-05T05:12:55Z","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":"ddd6936e5c6523ab1c622a20dfcf1e0538498f7a0ec3b71ac7e4be83ad23f93b","cross_cats_sorted":["cs.AI","cs.LG","cs.SI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2022-11-01T11:01:36Z","title_canon_sha256":"e751a54ed1818d160daa31bc65b714102bfef1d0fa479b91e7facd14dda7ece2"},"schema_version":"1.0","source":{"id":"2211.00384","kind":"arxiv","version":2}},"canonical_sha256":"00b13b0c044fab03a50ffcdd76bbe8fa7f969f88a6c9a9a4fffabd4b0cee0465","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"00b13b0c044fab03a50ffcdd76bbe8fa7f969f88a6c9a9a4fffabd4b0cee0465","first_computed_at":"2026-07-05T05:12:55.149304Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T05:12:55.149304Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"dL1zvpV2kvw0A4tdKB4xEm1ciAGDz/CGYpd0R/L6nYthtxFsO7yPqx1K9Qv6Odt4ELhAu3Km/rRmveUosOG7Dg==","signature_status":"signed_v1","signed_at":"2026-07-05T05:12:55.149722Z","signed_message":"canonical_sha256_bytes"},"source_id":"2211.00384","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:94bf25adae03b592ed71e92a97f1b28db63c00c3fe913a778e6dbd4aa5603e29","sha256:8ecf8ba3a8952fb211a4a8a641b2ff927035a5e9c8b0ab1877c97dfb7d486e08"],"state_sha256":"d6b2f1eaea5718dde59f075f2bd5af3819f2609f01a28d58d97141f5fb3bb4f1"}