{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:NFZKTXUOIISUEYNLVS2X7DKTAU","short_pith_number":"pith:NFZKTXUO","canonical_record":{"source":{"id":"2607.19608","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-07-21T22:24:18Z","cross_cats_sorted":[],"title_canon_sha256":"49aafb2256d3df333176fdedbf8ba0d62b8e613b7f19df1dfa459b5f999c3535","abstract_canon_sha256":"0e32d764d3bf1258322871fedde2377f32180116935a465b8c58333aa928191a"},"schema_version":"1.0"},"canonical_sha256":"6972a9de8e42254261abacb57f8d5305379c8793543b7246d8d005c87b98c726","source":{"kind":"arxiv","id":"2607.19608","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2607.19608","created_at":"2026-07-23T00:23:59Z"},{"alias_kind":"arxiv_version","alias_value":"2607.19608v1","created_at":"2026-07-23T00:23:59Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.19608","created_at":"2026-07-23T00:23:59Z"},{"alias_kind":"pith_short_12","alias_value":"NFZKTXUOIISU","created_at":"2026-07-23T00:23:59Z"},{"alias_kind":"pith_short_16","alias_value":"NFZKTXUOIISUEYNL","created_at":"2026-07-23T00:23:59Z"},{"alias_kind":"pith_short_8","alias_value":"NFZKTXUO","created_at":"2026-07-23T00:23:59Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:NFZKTXUOIISUEYNLVS2X7DKTAU","target":"record","payload":{"canonical_record":{"source":{"id":"2607.19608","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-07-21T22:24:18Z","cross_cats_sorted":[],"title_canon_sha256":"49aafb2256d3df333176fdedbf8ba0d62b8e613b7f19df1dfa459b5f999c3535","abstract_canon_sha256":"0e32d764d3bf1258322871fedde2377f32180116935a465b8c58333aa928191a"},"schema_version":"1.0"},"canonical_sha256":"6972a9de8e42254261abacb57f8d5305379c8793543b7246d8d005c87b98c726","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-23T00:23:59.050290Z","signature_b64":"+CyIdjBa7VOuPC/11GFlSnp/ffNuWbj+WU5dzgwSTGBreUQm8YQwL2bMxpzqLTUbFkj7wp8y92OhVQAg9o6FCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6972a9de8e42254261abacb57f8d5305379c8793543b7246d8d005c87b98c726","last_reissued_at":"2026-07-23T00:23:59.049461Z","signature_status":"signed_v1","first_computed_at":"2026-07-23T00:23:59.049461Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2607.19608","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":"2026-07-23T00:23:59Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"1bQ5Ph7w8GLcBdpPUjfR5+YqWsamdRuhMJIhVt2Dmxdm9VKced+yCzKsVKfmcIFWTAVJqCsLcdUPqnsaED2bCg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-15T05:14:45.627857Z"},"content_sha256":"94c49ebbab001fd1c78187ad96659038a9ede501cbbadd81f423326947cecafa","schema_version":"1.0","event_id":"sha256:94c49ebbab001fd1c78187ad96659038a9ede501cbbadd81f423326947cecafa"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:NFZKTXUOIISUEYNLVS2X7DKTAU","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Task Competence Is Not Instruction Following: Evaluating Instruction-Conflicting Behavior in Small Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Mahdiyeh Farajidizaji, Vatsal Raina","submitted_at":"2026-07-21T22:24:18Z","abstract_excerpt":"Instruction tuning is meant to make language models follow user requests, yet it is unclear whether small models comply when an instruction conflicts with their usual task behavior. We study this across three tasks - multiple-choice question answering (MCQA), sentiment classification, and mathematical question answering - by pairing a standard instruction with a conflicting non-standard one (select an incorrect option, output the opposite sentiment, or return twice the answer). This cross-task design allows us to test whether resistance to conflicting instructions is tied to specific task char"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.19608","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/2607.19608/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":"2026-07-23T00:23:59Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"D8kCyHwfy8723cPLeyUorlNLC556GqshqM4y6aYskzPkvtRLMq37Y+URspp4T5vDA2bvB7N6MV0+N8e2efc2Bw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-15T05:14:45.628377Z"},"content_sha256":"f291d787b61d19e162c89a5fb98aa81e736e83ee4989dbef39286af7e746c7c5","schema_version":"1.0","event_id":"sha256:f291d787b61d19e162c89a5fb98aa81e736e83ee4989dbef39286af7e746c7c5"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/NFZKTXUOIISUEYNLVS2X7DKTAU/bundle.json","state_url":"https://pith.science/pith/NFZKTXUOIISUEYNLVS2X7DKTAU/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/NFZKTXUOIISUEYNLVS2X7DKTAU/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":"2026-08-15T05:14:45Z","links":{"resolver":"https://pith.science/pith/NFZKTXUOIISUEYNLVS2X7DKTAU","bundle":"https://pith.science/pith/NFZKTXUOIISUEYNLVS2X7DKTAU/bundle.json","state":"https://pith.science/pith/NFZKTXUOIISUEYNLVS2X7DKTAU/state.json","well_known_bundle":"https://pith.science/.well-known/pith/NFZKTXUOIISUEYNLVS2X7DKTAU/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:NFZKTXUOIISUEYNLVS2X7DKTAU","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":"0e32d764d3bf1258322871fedde2377f32180116935a465b8c58333aa928191a","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-07-21T22:24:18Z","title_canon_sha256":"49aafb2256d3df333176fdedbf8ba0d62b8e613b7f19df1dfa459b5f999c3535"},"schema_version":"1.0","source":{"id":"2607.19608","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2607.19608","created_at":"2026-07-23T00:23:59Z"},{"alias_kind":"arxiv_version","alias_value":"2607.19608v1","created_at":"2026-07-23T00:23:59Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.19608","created_at":"2026-07-23T00:23:59Z"},{"alias_kind":"pith_short_12","alias_value":"NFZKTXUOIISU","created_at":"2026-07-23T00:23:59Z"},{"alias_kind":"pith_short_16","alias_value":"NFZKTXUOIISUEYNL","created_at":"2026-07-23T00:23:59Z"},{"alias_kind":"pith_short_8","alias_value":"NFZKTXUO","created_at":"2026-07-23T00:23:59Z"}],"graph_snapshots":[{"event_id":"sha256:f291d787b61d19e162c89a5fb98aa81e736e83ee4989dbef39286af7e746c7c5","target":"graph","created_at":"2026-07-23T00:23:59Z","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/2607.19608/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Instruction tuning is meant to make language models follow user requests, yet it is unclear whether small models comply when an instruction conflicts with their usual task behavior. We study this across three tasks - multiple-choice question answering (MCQA), sentiment classification, and mathematical question answering - by pairing a standard instruction with a conflicting non-standard one (select an incorrect option, output the opposite sentiment, or return twice the answer). This cross-task design allows us to test whether resistance to conflicting instructions is tied to specific task char","authors_text":"Mahdiyeh Farajidizaji, Vatsal Raina","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-07-21T22:24:18Z","title":"Task Competence Is Not Instruction Following: Evaluating Instruction-Conflicting Behavior in Small Language Models"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.19608","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":"sha256:94c49ebbab001fd1c78187ad96659038a9ede501cbbadd81f423326947cecafa","target":"record","created_at":"2026-07-23T00:23:59Z","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":"0e32d764d3bf1258322871fedde2377f32180116935a465b8c58333aa928191a","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-07-21T22:24:18Z","title_canon_sha256":"49aafb2256d3df333176fdedbf8ba0d62b8e613b7f19df1dfa459b5f999c3535"},"schema_version":"1.0","source":{"id":"2607.19608","kind":"arxiv","version":1}},"canonical_sha256":"6972a9de8e42254261abacb57f8d5305379c8793543b7246d8d005c87b98c726","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"6972a9de8e42254261abacb57f8d5305379c8793543b7246d8d005c87b98c726","first_computed_at":"2026-07-23T00:23:59.049461Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-23T00:23:59.049461Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"+CyIdjBa7VOuPC/11GFlSnp/ffNuWbj+WU5dzgwSTGBreUQm8YQwL2bMxpzqLTUbFkj7wp8y92OhVQAg9o6FCA==","signature_status":"signed_v1","signed_at":"2026-07-23T00:23:59.050290Z","signed_message":"canonical_sha256_bytes"},"source_id":"2607.19608","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:94c49ebbab001fd1c78187ad96659038a9ede501cbbadd81f423326947cecafa","sha256:f291d787b61d19e162c89a5fb98aa81e736e83ee4989dbef39286af7e746c7c5"],"state_sha256":"85d5de61eed54a7f12cf5e557011f5ce6c5618c8b3ffb842e7d208cc6478cfa4"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"2yIYA1U+iWsZMCV8lyQDRKe0ZfrH06/neu8ta1sOD7qHEX4e5RtrXSSOl0SCPnXFBxIfpW9f/G9zQVQQ44kRAA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-15T05:14:45.633609Z","bundle_sha256":"543e020bb7caa04f2debeaff088c304a9763a6ce27aa27bfd40ec0522a2bb3ab"}}