{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:F5HNDDIR6N5BD4ZOEEG227LIFZ","short_pith_number":"pith:F5HNDDIR","canonical_record":{"source":{"id":"2606.11643","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-06-10T04:07:41Z","cross_cats_sorted":[],"title_canon_sha256":"7d9b904341e12333b4cf5341a09893b4da1a0a0974201c8c03d2ace4b68ff730","abstract_canon_sha256":"c947475b027a5fd9c5c4bf76f8c35e9c9a247a72ececda99f037d1b9ee3e6df2"},"schema_version":"1.0"},"canonical_sha256":"2f4ed18d11f37a11f32e210dad7d682e75d743c9b76374dfb4586dcf32b71de8","source":{"kind":"arxiv","id":"2606.11643","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.11643","created_at":"2026-06-11T01:10:00Z"},{"alias_kind":"arxiv_version","alias_value":"2606.11643v1","created_at":"2026-06-11T01:10:00Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.11643","created_at":"2026-06-11T01:10:00Z"},{"alias_kind":"pith_short_12","alias_value":"F5HNDDIR6N5B","created_at":"2026-06-11T01:10:00Z"},{"alias_kind":"pith_short_16","alias_value":"F5HNDDIR6N5BD4ZO","created_at":"2026-06-11T01:10:00Z"},{"alias_kind":"pith_short_8","alias_value":"F5HNDDIR","created_at":"2026-06-11T01:10:00Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:F5HNDDIR6N5BD4ZOEEG227LIFZ","target":"record","payload":{"canonical_record":{"source":{"id":"2606.11643","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-06-10T04:07:41Z","cross_cats_sorted":[],"title_canon_sha256":"7d9b904341e12333b4cf5341a09893b4da1a0a0974201c8c03d2ace4b68ff730","abstract_canon_sha256":"c947475b027a5fd9c5c4bf76f8c35e9c9a247a72ececda99f037d1b9ee3e6df2"},"schema_version":"1.0"},"canonical_sha256":"2f4ed18d11f37a11f32e210dad7d682e75d743c9b76374dfb4586dcf32b71de8","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-11T01:10:00.543259Z","signature_b64":"cUoErVAwZMDRWbYQXqxQZGo2KsXdshCqzdLkWY/F/RhcepGpTEj4Bu3UoM3kC1At4jNGqhyWWyJFpHE92qZxDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2f4ed18d11f37a11f32e210dad7d682e75d743c9b76374dfb4586dcf32b71de8","last_reissued_at":"2026-06-11T01:10:00.542532Z","signature_status":"signed_v1","first_computed_at":"2026-06-11T01:10:00.542532Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2606.11643","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":"2026-06-11T01:10:00Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"2iuZF+vszuZ8x4P0zb9z2NFjzXChUhT1JQ2ieL0EAh8uG4ixWrVDE5H0tG4a5V86uFZZs79Wn08hXxze4AVJDQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-29T23:08:44.091145Z"},"content_sha256":"3f328d52fbf687d5866bc4fe64c41393c49e5350f94cc8eaebf3e5f3ec498671","schema_version":"1.0","event_id":"sha256:3f328d52fbf687d5866bc4fe64c41393c49e5350f94cc8eaebf3e5f3ec498671"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:F5HNDDIR6N5BD4ZOEEG227LIFZ","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Improving Cross-Format Robustness in Language Models with Multi-Format Training","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Dingnan Jin, He Cao, June M. Liu, Jun Zhou, Qing Cui, Shaomian Zheng","submitted_at":"2026-06-10T04:07:41Z","abstract_excerpt":"Large language models often remain sensitive to answer format: a question solved correctly in one form may fail in another semantically equivalent form. To study this gap, we define cross-format robustness as the extent to which a model answers the same underlying question consistently across formats. We then compare full-format training with FormatMix, which expands only a subset of training items into multiple equivalent formats using either random or targeted selection. Across GLM4 and Llama-3.1, multi-format supervision consistently improves both task performance and cross-format robustnes"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.11643","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/2606.11643/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":"2026-06-11T01:10:00Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"DYDZzCg0ddqveJCocZd8adNejwf9f6U/S0ow/I8/vlI6M1yPKvFcUkRUdfJLVlCppoX6JBP2y+mQJbfYweQbCw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-29T23:08:44.091529Z"},"content_sha256":"2a7a0a9ba3ece70a15583a532c5440a719760606df599eb58c91b1ddda8c5c6e","schema_version":"1.0","event_id":"sha256:2a7a0a9ba3ece70a15583a532c5440a719760606df599eb58c91b1ddda8c5c6e"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/F5HNDDIR6N5BD4ZOEEG227LIFZ/bundle.json","state_url":"https://pith.science/pith/F5HNDDIR6N5BD4ZOEEG227LIFZ/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/F5HNDDIR6N5BD4ZOEEG227LIFZ/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":"2026-06-29T23:08:44Z","links":{"resolver":"https://pith.science/pith/F5HNDDIR6N5BD4ZOEEG227LIFZ","bundle":"https://pith.science/pith/F5HNDDIR6N5BD4ZOEEG227LIFZ/bundle.json","state":"https://pith.science/pith/F5HNDDIR6N5BD4ZOEEG227LIFZ/state.json","well_known_bundle":"https://pith.science/.well-known/pith/F5HNDDIR6N5BD4ZOEEG227LIFZ/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:F5HNDDIR6N5BD4ZOEEG227LIFZ","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":"c947475b027a5fd9c5c4bf76f8c35e9c9a247a72ececda99f037d1b9ee3e6df2","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-06-10T04:07:41Z","title_canon_sha256":"7d9b904341e12333b4cf5341a09893b4da1a0a0974201c8c03d2ace4b68ff730"},"schema_version":"1.0","source":{"id":"2606.11643","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.11643","created_at":"2026-06-11T01:10:00Z"},{"alias_kind":"arxiv_version","alias_value":"2606.11643v1","created_at":"2026-06-11T01:10:00Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.11643","created_at":"2026-06-11T01:10:00Z"},{"alias_kind":"pith_short_12","alias_value":"F5HNDDIR6N5B","created_at":"2026-06-11T01:10:00Z"},{"alias_kind":"pith_short_16","alias_value":"F5HNDDIR6N5BD4ZO","created_at":"2026-06-11T01:10:00Z"},{"alias_kind":"pith_short_8","alias_value":"F5HNDDIR","created_at":"2026-06-11T01:10:00Z"}],"graph_snapshots":[{"event_id":"sha256:2a7a0a9ba3ece70a15583a532c5440a719760606df599eb58c91b1ddda8c5c6e","target":"graph","created_at":"2026-06-11T01:10:00Z","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/2606.11643/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large language models often remain sensitive to answer format: a question solved correctly in one form may fail in another semantically equivalent form. To study this gap, we define cross-format robustness as the extent to which a model answers the same underlying question consistently across formats. We then compare full-format training with FormatMix, which expands only a subset of training items into multiple equivalent formats using either random or targeted selection. Across GLM4 and Llama-3.1, multi-format supervision consistently improves both task performance and cross-format robustnes","authors_text":"Dingnan Jin, He Cao, June M. Liu, Jun Zhou, Qing Cui, Shaomian Zheng","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-06-10T04:07:41Z","title":"Improving Cross-Format Robustness in Language Models with Multi-Format Training"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.11643","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":"sha256:3f328d52fbf687d5866bc4fe64c41393c49e5350f94cc8eaebf3e5f3ec498671","target":"record","created_at":"2026-06-11T01:10:00Z","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":"c947475b027a5fd9c5c4bf76f8c35e9c9a247a72ececda99f037d1b9ee3e6df2","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-06-10T04:07:41Z","title_canon_sha256":"7d9b904341e12333b4cf5341a09893b4da1a0a0974201c8c03d2ace4b68ff730"},"schema_version":"1.0","source":{"id":"2606.11643","kind":"arxiv","version":1}},"canonical_sha256":"2f4ed18d11f37a11f32e210dad7d682e75d743c9b76374dfb4586dcf32b71de8","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"2f4ed18d11f37a11f32e210dad7d682e75d743c9b76374dfb4586dcf32b71de8","first_computed_at":"2026-06-11T01:10:00.542532Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-11T01:10:00.542532Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"cUoErVAwZMDRWbYQXqxQZGo2KsXdshCqzdLkWY/F/RhcepGpTEj4Bu3UoM3kC1At4jNGqhyWWyJFpHE92qZxDA==","signature_status":"signed_v1","signed_at":"2026-06-11T01:10:00.543259Z","signed_message":"canonical_sha256_bytes"},"source_id":"2606.11643","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:3f328d52fbf687d5866bc4fe64c41393c49e5350f94cc8eaebf3e5f3ec498671","sha256:2a7a0a9ba3ece70a15583a532c5440a719760606df599eb58c91b1ddda8c5c6e"],"state_sha256":"01e3622868849fdd2ef077c33f2a29a9134f2734bca2b117db16d5ad017efb6a"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"rKCUICt7eYywlvbJPEd0yOjFQxyqj/VJX+DOiw21p7578AEa6ynjmztavGrWT+8EVSCjrPn8TjEDjvcdnsYLBQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-29T23:08:44.093495Z","bundle_sha256":"cc504fdc80ba1b2f773a6f553cd19968acf89cc2335ae065f7a4f1783ac890ad"}}