{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2021:T4KW4UGFFBYC7CRJDDM777FKYL","short_pith_number":"pith:T4KW4UGF","canonical_record":{"source":{"id":"2101.04840","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2021-01-13T02:37:54Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"47f9fecc24d696620042b1946a29a44c4630701c50fd305f9f2602b05694460d","abstract_canon_sha256":"d2dd40f4066f5779913d7b3a61134e0f43c244903be7e208017e46fa0f4fddda"},"schema_version":"1.0"},"canonical_sha256":"9f156e50c528702f8a2918d9fffcaac2f3deb029935b064d70919cd6818607e1","source":{"kind":"arxiv","id":"2101.04840","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2101.04840","created_at":"2026-07-05T02:06:35Z"},{"alias_kind":"arxiv_version","alias_value":"2101.04840v1","created_at":"2026-07-05T02:06:35Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2101.04840","created_at":"2026-07-05T02:06:35Z"},{"alias_kind":"pith_short_12","alias_value":"T4KW4UGFFBYC","created_at":"2026-07-05T02:06:35Z"},{"alias_kind":"pith_short_16","alias_value":"T4KW4UGFFBYC7CRJ","created_at":"2026-07-05T02:06:35Z"},{"alias_kind":"pith_short_8","alias_value":"T4KW4UGF","created_at":"2026-07-05T02:06:35Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2021:T4KW4UGFFBYC7CRJDDM777FKYL","target":"record","payload":{"canonical_record":{"source":{"id":"2101.04840","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2021-01-13T02:37:54Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"47f9fecc24d696620042b1946a29a44c4630701c50fd305f9f2602b05694460d","abstract_canon_sha256":"d2dd40f4066f5779913d7b3a61134e0f43c244903be7e208017e46fa0f4fddda"},"schema_version":"1.0"},"canonical_sha256":"9f156e50c528702f8a2918d9fffcaac2f3deb029935b064d70919cd6818607e1","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:06:35.620881Z","signature_b64":"baalJTel2HKRrsRZgyZ9LXplB2664rJhT8GfkiuTTgwdw1YJSHy2SSM5b+b7UG+oUx58wIyu3xHx40yixH15Cg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9f156e50c528702f8a2918d9fffcaac2f3deb029935b064d70919cd6818607e1","last_reissued_at":"2026-07-05T02:06:35.620471Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:06:35.620471Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2101.04840","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":"2026-07-05T02:06:35Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"hhlXNTfIsal0IHqs7ZFIaaft5INlgY7d5STTqcJm9eES7hMcqQ9EDIky9ZRL1uH+/F88I0gpjGE6yBuUmn+oBQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-06T09:11:25.433548Z"},"content_sha256":"3342e603c6c13f8f4a1958f22d565d9f715546cc98a6d7cb3f7b5f4d3a49d896","schema_version":"1.0","event_id":"sha256:3342e603c6c13f8f4a1958f22d565d9f715546cc98a6d7cb3f7b5f4d3a49d896"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2021:T4KW4UGFFBYC7CRJDDM777FKYL","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Robustness Gym: Unifying the NLP Evaluation Landscape","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Caiming Xiong, Christopher R\\'e, Jason Wu, Jesse Vig, Karan Goel, Mohit Bansal, Nazneen Rajani, Samson Tan, Stephan Zheng","submitted_at":"2021-01-13T02:37:54Z","abstract_excerpt":"Despite impressive performance on standard benchmarks, deep neural networks are often brittle when deployed in real-world systems. Consequently, recent research has focused on testing the robustness of such models, resulting in a diverse set of evaluation methodologies ranging from adversarial attacks to rule-based data transformations. In this work, we identify challenges with evaluating NLP systems and propose a solution in the form of Robustness Gym (RG), a simple and extensible evaluation toolkit that unifies 4 standard evaluation paradigms: subpopulations, transformations, evaluation sets"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2101.04840","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/2101.04840/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":"2026-07-05T02:06:35Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"/PJiDxBWjbVVXgvl2Bsihc05EuxNo+tvBgkY1g145BDeGNNVt/lHOcEZDrFtF2NpD8WiumH2UPmPx5dcJ29qCg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-06T09:11:25.434122Z"},"content_sha256":"91f9f1d28e21b161445bdb1c82d9e40625d468024b61df05103249dfe891d5c8","schema_version":"1.0","event_id":"sha256:91f9f1d28e21b161445bdb1c82d9e40625d468024b61df05103249dfe891d5c8"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/T4KW4UGFFBYC7CRJDDM777FKYL/bundle.json","state_url":"https://pith.science/pith/T4KW4UGFFBYC7CRJDDM777FKYL/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/T4KW4UGFFBYC7CRJDDM777FKYL/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":"2026-07-06T09:11:25Z","links":{"resolver":"https://pith.science/pith/T4KW4UGFFBYC7CRJDDM777FKYL","bundle":"https://pith.science/pith/T4KW4UGFFBYC7CRJDDM777FKYL/bundle.json","state":"https://pith.science/pith/T4KW4UGFFBYC7CRJDDM777FKYL/state.json","well_known_bundle":"https://pith.science/.well-known/pith/T4KW4UGFFBYC7CRJDDM777FKYL/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2021:T4KW4UGFFBYC7CRJDDM777FKYL","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":"d2dd40f4066f5779913d7b3a61134e0f43c244903be7e208017e46fa0f4fddda","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2021-01-13T02:37:54Z","title_canon_sha256":"47f9fecc24d696620042b1946a29a44c4630701c50fd305f9f2602b05694460d"},"schema_version":"1.0","source":{"id":"2101.04840","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2101.04840","created_at":"2026-07-05T02:06:35Z"},{"alias_kind":"arxiv_version","alias_value":"2101.04840v1","created_at":"2026-07-05T02:06:35Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2101.04840","created_at":"2026-07-05T02:06:35Z"},{"alias_kind":"pith_short_12","alias_value":"T4KW4UGFFBYC","created_at":"2026-07-05T02:06:35Z"},{"alias_kind":"pith_short_16","alias_value":"T4KW4UGFFBYC7CRJ","created_at":"2026-07-05T02:06:35Z"},{"alias_kind":"pith_short_8","alias_value":"T4KW4UGF","created_at":"2026-07-05T02:06:35Z"}],"graph_snapshots":[{"event_id":"sha256:91f9f1d28e21b161445bdb1c82d9e40625d468024b61df05103249dfe891d5c8","target":"graph","created_at":"2026-07-05T02:06:35Z","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/2101.04840/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Despite impressive performance on standard benchmarks, deep neural networks are often brittle when deployed in real-world systems. Consequently, recent research has focused on testing the robustness of such models, resulting in a diverse set of evaluation methodologies ranging from adversarial attacks to rule-based data transformations. In this work, we identify challenges with evaluating NLP systems and propose a solution in the form of Robustness Gym (RG), a simple and extensible evaluation toolkit that unifies 4 standard evaluation paradigms: subpopulations, transformations, evaluation sets","authors_text":"Caiming Xiong, Christopher R\\'e, Jason Wu, Jesse Vig, Karan Goel, Mohit Bansal, Nazneen Rajani, Samson Tan, Stephan Zheng","cross_cats":["cs.AI","cs.LG"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2021-01-13T02:37:54Z","title":"Robustness Gym: Unifying the NLP Evaluation Landscape"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2101.04840","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":"sha256:3342e603c6c13f8f4a1958f22d565d9f715546cc98a6d7cb3f7b5f4d3a49d896","target":"record","created_at":"2026-07-05T02:06:35Z","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":"d2dd40f4066f5779913d7b3a61134e0f43c244903be7e208017e46fa0f4fddda","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2021-01-13T02:37:54Z","title_canon_sha256":"47f9fecc24d696620042b1946a29a44c4630701c50fd305f9f2602b05694460d"},"schema_version":"1.0","source":{"id":"2101.04840","kind":"arxiv","version":1}},"canonical_sha256":"9f156e50c528702f8a2918d9fffcaac2f3deb029935b064d70919cd6818607e1","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"9f156e50c528702f8a2918d9fffcaac2f3deb029935b064d70919cd6818607e1","first_computed_at":"2026-07-05T02:06:35.620471Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T02:06:35.620471Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"baalJTel2HKRrsRZgyZ9LXplB2664rJhT8GfkiuTTgwdw1YJSHy2SSM5b+b7UG+oUx58wIyu3xHx40yixH15Cg==","signature_status":"signed_v1","signed_at":"2026-07-05T02:06:35.620881Z","signed_message":"canonical_sha256_bytes"},"source_id":"2101.04840","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:3342e603c6c13f8f4a1958f22d565d9f715546cc98a6d7cb3f7b5f4d3a49d896","sha256:91f9f1d28e21b161445bdb1c82d9e40625d468024b61df05103249dfe891d5c8"],"state_sha256":"1bb42c3a9b0ff51ba0f9dd3cffca7815c8f3b4c7848f585536508e60a8b1aaf8"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ef01rezI8UilyQU/93ZVEf3eursr9MXqrmGD/KlP8oSm6f5WSbFD+Nswsm8lhbq6a67iuP/TTdm2DGrOpGebBw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-06T09:11:25.436154Z","bundle_sha256":"35c6ceb693cd5255b9333a923f57aabcdaaf686f309817c3ace270a8a696eb21"}}