{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2022:HLKOVFTTVGBNIDNOAUXFYCH7PL","short_pith_number":"pith:HLKOVFTT","canonical_record":{"source":{"id":"2205.12600","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2022-05-25T09:25:06Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"499cb6159360003c29f3660de20e50973827e18da54237aaf8ab60d3ae8600b5","abstract_canon_sha256":"586a40fba258ad31d941557cdcf1aac881b879a57a4428bffa676253bb306b2d"},"schema_version":"1.0"},"canonical_sha256":"3ad4ea9673a982d40dae052e5c08ff7af74969f2934e70b78d4b682d58a957c6","source":{"kind":"arxiv","id":"2205.12600","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2205.12600","created_at":"2026-07-05T04:26:27Z"},{"alias_kind":"arxiv_version","alias_value":"2205.12600v1","created_at":"2026-07-05T04:26:27Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2205.12600","created_at":"2026-07-05T04:26:27Z"},{"alias_kind":"pith_short_12","alias_value":"HLKOVFTTVGBN","created_at":"2026-07-05T04:26:27Z"},{"alias_kind":"pith_short_16","alias_value":"HLKOVFTTVGBNIDNO","created_at":"2026-07-05T04:26:27Z"},{"alias_kind":"pith_short_8","alias_value":"HLKOVFTT","created_at":"2026-07-05T04:26:27Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2022:HLKOVFTTVGBNIDNOAUXFYCH7PL","target":"record","payload":{"canonical_record":{"source":{"id":"2205.12600","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2022-05-25T09:25:06Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"499cb6159360003c29f3660de20e50973827e18da54237aaf8ab60d3ae8600b5","abstract_canon_sha256":"586a40fba258ad31d941557cdcf1aac881b879a57a4428bffa676253bb306b2d"},"schema_version":"1.0"},"canonical_sha256":"3ad4ea9673a982d40dae052e5c08ff7af74969f2934e70b78d4b682d58a957c6","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T04:26:27.759309Z","signature_b64":"35OSRvbuRdpOkE/08Kh9/L3l8dbbZzNzr0GzKOqwfQK7W5bYY6D5kqqPmDFARZyMrh74V4r5C3XzEngpnrMbDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3ad4ea9673a982d40dae052e5c08ff7af74969f2934e70b78d4b682d58a957c6","last_reissued_at":"2026-07-05T04:26:27.758895Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T04:26:27.758895Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2205.12600","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T04:26:27Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"548MDYqckswrZzpF3RhexZf9L0iMI+v0HK7itCjcUy2YctFkArhsd6RNFus7Z4FRBLVgMOHZVvx7wKeXpnj8AQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-17T17:44:56.903126Z"},"content_sha256":"287bffe1ad31ddc69dafb11f8e6e9c330ec6979d10823f20f3d8bd7953e68408","schema_version":"1.0","event_id":"sha256:287bffe1ad31ddc69dafb11f8e6e9c330ec6979d10823f20f3d8bd7953e68408"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2022:HLKOVFTTVGBNIDNOAUXFYCH7PL","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"ORCA: Interpreting Prompted Language Models via Locating Supporting Data Evidence in the Ocean of Pretraining Data","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Xiaochuang Han, Yulia Tsvetkov","submitted_at":"2022-05-25T09:25:06Z","abstract_excerpt":"Large pretrained language models have been performing increasingly well in a variety of downstream tasks via prompting. However, it remains unclear from where the model learns the task-specific knowledge, especially in a zero-shot setup. In this work, we want to find evidence of the model's task-specific competence from pretraining and are specifically interested in locating a very small subset of pretraining data that directly supports the model in the task. We call such a subset supporting data evidence and propose a novel method ORCA to effectively identify it, by iteratively using gradient"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2205.12600","kind":"arxiv","version":1},"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/2205.12600/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":"2026-07-05T04:26:27Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"0pJ5DcaAFONyi1l5OB21saCW1HXeSD3g9j7kLqLoeaVC2RMzC+VId0tjbkiAeD74qJz82QnVHp9JoKzaH9oXBA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-17T17:44:56.903664Z"},"content_sha256":"d36c9fad7656b38690f99d6f51331744775610039856e82f600f04eacb57b565","schema_version":"1.0","event_id":"sha256:d36c9fad7656b38690f99d6f51331744775610039856e82f600f04eacb57b565"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/HLKOVFTTVGBNIDNOAUXFYCH7PL/bundle.json","state_url":"https://pith.science/pith/HLKOVFTTVGBNIDNOAUXFYCH7PL/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/HLKOVFTTVGBNIDNOAUXFYCH7PL/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":"2026-08-17T17:44:56Z","links":{"resolver":"https://pith.science/pith/HLKOVFTTVGBNIDNOAUXFYCH7PL","bundle":"https://pith.science/pith/HLKOVFTTVGBNIDNOAUXFYCH7PL/bundle.json","state":"https://pith.science/pith/HLKOVFTTVGBNIDNOAUXFYCH7PL/state.json","well_known_bundle":"https://pith.science/.well-known/pith/HLKOVFTTVGBNIDNOAUXFYCH7PL/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2022:HLKOVFTTVGBNIDNOAUXFYCH7PL","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":"586a40fba258ad31d941557cdcf1aac881b879a57a4428bffa676253bb306b2d","cross_cats_sorted":["cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2022-05-25T09:25:06Z","title_canon_sha256":"499cb6159360003c29f3660de20e50973827e18da54237aaf8ab60d3ae8600b5"},"schema_version":"1.0","source":{"id":"2205.12600","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2205.12600","created_at":"2026-07-05T04:26:27Z"},{"alias_kind":"arxiv_version","alias_value":"2205.12600v1","created_at":"2026-07-05T04:26:27Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2205.12600","created_at":"2026-07-05T04:26:27Z"},{"alias_kind":"pith_short_12","alias_value":"HLKOVFTTVGBN","created_at":"2026-07-05T04:26:27Z"},{"alias_kind":"pith_short_16","alias_value":"HLKOVFTTVGBNIDNO","created_at":"2026-07-05T04:26:27Z"},{"alias_kind":"pith_short_8","alias_value":"HLKOVFTT","created_at":"2026-07-05T04:26:27Z"}],"graph_snapshots":[{"event_id":"sha256:d36c9fad7656b38690f99d6f51331744775610039856e82f600f04eacb57b565","target":"graph","created_at":"2026-07-05T04:26:27Z","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/2205.12600/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large pretrained language models have been performing increasingly well in a variety of downstream tasks via prompting. However, it remains unclear from where the model learns the task-specific knowledge, especially in a zero-shot setup. In this work, we want to find evidence of the model's task-specific competence from pretraining and are specifically interested in locating a very small subset of pretraining data that directly supports the model in the task. We call such a subset supporting data evidence and propose a novel method ORCA to effectively identify it, by iteratively using gradient","authors_text":"Xiaochuang Han, Yulia Tsvetkov","cross_cats":["cs.LG"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2022-05-25T09:25:06Z","title":"ORCA: Interpreting Prompted Language Models via Locating Supporting Data Evidence in the Ocean of Pretraining Data"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2205.12600","kind":"arxiv","version":1},"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":"sha256:287bffe1ad31ddc69dafb11f8e6e9c330ec6979d10823f20f3d8bd7953e68408","target":"record","created_at":"2026-07-05T04:26:27Z","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":"586a40fba258ad31d941557cdcf1aac881b879a57a4428bffa676253bb306b2d","cross_cats_sorted":["cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2022-05-25T09:25:06Z","title_canon_sha256":"499cb6159360003c29f3660de20e50973827e18da54237aaf8ab60d3ae8600b5"},"schema_version":"1.0","source":{"id":"2205.12600","kind":"arxiv","version":1}},"canonical_sha256":"3ad4ea9673a982d40dae052e5c08ff7af74969f2934e70b78d4b682d58a957c6","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"3ad4ea9673a982d40dae052e5c08ff7af74969f2934e70b78d4b682d58a957c6","first_computed_at":"2026-07-05T04:26:27.758895Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T04:26:27.758895Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"35OSRvbuRdpOkE/08Kh9/L3l8dbbZzNzr0GzKOqwfQK7W5bYY6D5kqqPmDFARZyMrh74V4r5C3XzEngpnrMbDw==","signature_status":"signed_v1","signed_at":"2026-07-05T04:26:27.759309Z","signed_message":"canonical_sha256_bytes"},"source_id":"2205.12600","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:287bffe1ad31ddc69dafb11f8e6e9c330ec6979d10823f20f3d8bd7953e68408","sha256:d36c9fad7656b38690f99d6f51331744775610039856e82f600f04eacb57b565"],"state_sha256":"1b4f679e9857e79dcb06d6f89a277a5e477ffd2eb37ba8fcc2ee3d31bf60b33c"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"dzb/n1Xi8Xeahs00yVqU9U3M8tvDAHqHF6U3eMivgUATyEMDFpAdRtn3RcNFC/+iI3hy8ZngZqfpnsOOV9qEBg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-17T17:44:56.907533Z","bundle_sha256":"2a1c3e1e9e5f36a69c46981092fc95bc689493764e649240ebbfd2f31dc0f7f5"}}