{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2020:3H2GJRDX565NEJJILYMNNMCASU","short_pith_number":"pith:3H2GJRDX","canonical_record":{"source":{"id":"2012.14666","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2020-12-29T08:58:49Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"e2b8a178ccdbfffbe7352223fcead90d5ff8ea5dc1cb9144a12e7650d44a484f","abstract_canon_sha256":"daaebfa1fcd80f4001980eb117648f62bb0221c90c0e4183d3a9e5812f18d2e9"},"schema_version":"1.0"},"canonical_sha256":"d9f464c477efbad225285e18d6b040952ccdff8935e4b0b1f98225163289fa33","source":{"kind":"arxiv","id":"2012.14666","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2012.14666","created_at":"2026-07-05T02:03:00Z"},{"alias_kind":"arxiv_version","alias_value":"2012.14666v1","created_at":"2026-07-05T02:03:00Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2012.14666","created_at":"2026-07-05T02:03:00Z"},{"alias_kind":"pith_short_12","alias_value":"3H2GJRDX565N","created_at":"2026-07-05T02:03:00Z"},{"alias_kind":"pith_short_16","alias_value":"3H2GJRDX565NEJJI","created_at":"2026-07-05T02:03:00Z"},{"alias_kind":"pith_short_8","alias_value":"3H2GJRDX","created_at":"2026-07-05T02:03:00Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2020:3H2GJRDX565NEJJILYMNNMCASU","target":"record","payload":{"canonical_record":{"source":{"id":"2012.14666","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2020-12-29T08:58:49Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"e2b8a178ccdbfffbe7352223fcead90d5ff8ea5dc1cb9144a12e7650d44a484f","abstract_canon_sha256":"daaebfa1fcd80f4001980eb117648f62bb0221c90c0e4183d3a9e5812f18d2e9"},"schema_version":"1.0"},"canonical_sha256":"d9f464c477efbad225285e18d6b040952ccdff8935e4b0b1f98225163289fa33","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:03:00.462099Z","signature_b64":"Ug9vJuBoKC7Dyuw0yxKxQCQcPWOFFzy533wjIrkh1SwWdES8pzDZXBah+BPvR52x/t2q3gNdGm6fNtCElB3xDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d9f464c477efbad225285e18d6b040952ccdff8935e4b0b1f98225163289fa33","last_reissued_at":"2026-07-05T02:03:00.461694Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:03:00.461694Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2012.14666","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":"2026-07-05T02:03:00Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"TkJjHI3yDaH+BofOwz2j2PXbTi6u9fagAULC5snKp+dNGew+yfagmKjDDlwUT1G3YOJWWPYxm/J06bl0Gu8KDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-03T17:33:55.033617Z"},"content_sha256":"67683463ab871f9f0d0d8d5b7ceea3060518170ce306c52e05965d5dc3b7ae8e","schema_version":"1.0","event_id":"sha256:67683463ab871f9f0d0d8d5b7ceea3060518170ce306c52e05965d5dc3b7ae8e"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2020:3H2GJRDX565NEJJILYMNNMCASU","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"RADDLE: An Evaluation Benchmark and Analysis Platform for Robust Task-oriented Dialog Systems","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Baolin Peng, Chenguang Zhu, Chunyuan Li, Jianfeng Gao, Jinchao Li, Zhu Zhang","submitted_at":"2020-12-29T08:58:49Z","abstract_excerpt":"For task-oriented dialog systems to be maximally useful, it must be able to process conversations in a way that is (1) generalizable with a small number of training examples for new task domains, and (2) robust to user input in various styles, modalities or domains. In pursuit of these goals, we introduce the RADDLE benchmark, a collection of corpora and tools for evaluating the performance of models across a diverse set of domains. By including tasks with limited training data, RADDLE is designed to favor and encourage models with a strong generalization ability. RADDLE also includes a diagno"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2012.14666","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/2012.14666/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":"2026-07-05T02:03:00Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"NVCPumXb+d4RxNeTAPcnyN0wLSMdT7JqPBz4qdLMjo1S0lx4r3bGY9paNNBo1Mpv0cdyS40sF7gX3XxYKf7PAw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-03T17:33:55.034504Z"},"content_sha256":"184fc14a99655e20a8b8e729e988f43604f752aacb4948b91942fc3ade54815a","schema_version":"1.0","event_id":"sha256:184fc14a99655e20a8b8e729e988f43604f752aacb4948b91942fc3ade54815a"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/3H2GJRDX565NEJJILYMNNMCASU/bundle.json","state_url":"https://pith.science/pith/3H2GJRDX565NEJJILYMNNMCASU/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/3H2GJRDX565NEJJILYMNNMCASU/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":"2026-08-03T17:33:55Z","links":{"resolver":"https://pith.science/pith/3H2GJRDX565NEJJILYMNNMCASU","bundle":"https://pith.science/pith/3H2GJRDX565NEJJILYMNNMCASU/bundle.json","state":"https://pith.science/pith/3H2GJRDX565NEJJILYMNNMCASU/state.json","well_known_bundle":"https://pith.science/.well-known/pith/3H2GJRDX565NEJJILYMNNMCASU/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2020:3H2GJRDX565NEJJILYMNNMCASU","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":"daaebfa1fcd80f4001980eb117648f62bb0221c90c0e4183d3a9e5812f18d2e9","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2020-12-29T08:58:49Z","title_canon_sha256":"e2b8a178ccdbfffbe7352223fcead90d5ff8ea5dc1cb9144a12e7650d44a484f"},"schema_version":"1.0","source":{"id":"2012.14666","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2012.14666","created_at":"2026-07-05T02:03:00Z"},{"alias_kind":"arxiv_version","alias_value":"2012.14666v1","created_at":"2026-07-05T02:03:00Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2012.14666","created_at":"2026-07-05T02:03:00Z"},{"alias_kind":"pith_short_12","alias_value":"3H2GJRDX565N","created_at":"2026-07-05T02:03:00Z"},{"alias_kind":"pith_short_16","alias_value":"3H2GJRDX565NEJJI","created_at":"2026-07-05T02:03:00Z"},{"alias_kind":"pith_short_8","alias_value":"3H2GJRDX","created_at":"2026-07-05T02:03:00Z"}],"graph_snapshots":[{"event_id":"sha256:184fc14a99655e20a8b8e729e988f43604f752aacb4948b91942fc3ade54815a","target":"graph","created_at":"2026-07-05T02:03:00Z","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/2012.14666/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"For task-oriented dialog systems to be maximally useful, it must be able to process conversations in a way that is (1) generalizable with a small number of training examples for new task domains, and (2) robust to user input in various styles, modalities or domains. In pursuit of these goals, we introduce the RADDLE benchmark, a collection of corpora and tools for evaluating the performance of models across a diverse set of domains. By including tasks with limited training data, RADDLE is designed to favor and encourage models with a strong generalization ability. RADDLE also includes a diagno","authors_text":"Baolin Peng, Chenguang Zhu, Chunyuan Li, Jianfeng Gao, Jinchao Li, Zhu Zhang","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2020-12-29T08:58:49Z","title":"RADDLE: An Evaluation Benchmark and Analysis Platform for Robust Task-oriented Dialog Systems"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2012.14666","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":"sha256:67683463ab871f9f0d0d8d5b7ceea3060518170ce306c52e05965d5dc3b7ae8e","target":"record","created_at":"2026-07-05T02:03:00Z","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":"daaebfa1fcd80f4001980eb117648f62bb0221c90c0e4183d3a9e5812f18d2e9","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2020-12-29T08:58:49Z","title_canon_sha256":"e2b8a178ccdbfffbe7352223fcead90d5ff8ea5dc1cb9144a12e7650d44a484f"},"schema_version":"1.0","source":{"id":"2012.14666","kind":"arxiv","version":1}},"canonical_sha256":"d9f464c477efbad225285e18d6b040952ccdff8935e4b0b1f98225163289fa33","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"d9f464c477efbad225285e18d6b040952ccdff8935e4b0b1f98225163289fa33","first_computed_at":"2026-07-05T02:03:00.461694Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T02:03:00.461694Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"Ug9vJuBoKC7Dyuw0yxKxQCQcPWOFFzy533wjIrkh1SwWdES8pzDZXBah+BPvR52x/t2q3gNdGm6fNtCElB3xDw==","signature_status":"signed_v1","signed_at":"2026-07-05T02:03:00.462099Z","signed_message":"canonical_sha256_bytes"},"source_id":"2012.14666","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:67683463ab871f9f0d0d8d5b7ceea3060518170ce306c52e05965d5dc3b7ae8e","sha256:184fc14a99655e20a8b8e729e988f43604f752aacb4948b91942fc3ade54815a"],"state_sha256":"8b5a4c7ed7c9932b62b70ce3662c40b91cee92086ec9458584f3f172bea856d5"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"iUZMzUxW5OBZO0qsVXsB8Zjko4mYjtDjeySr8FouS3E/AaIko/yEudWSiPFQ3BgU/jlILHr9T0nHQ1JLblmdAQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-03T17:33:55.041312Z","bundle_sha256":"098a862ce07d4274d28cebe96cf315ab24059a772351b518add696c97226ee2b"}}