{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:7P4MQPRZW4UKJLBA7LRVLSCG4K","short_pith_number":"pith:7P4MQPRZ","canonical_record":{"source":{"id":"2605.11632","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-05-12T06:56:18Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"a6e1694b654f559b8075b950f5cc4ba25911443e16acdafd56ee72a1e31a47f8","abstract_canon_sha256":"6061ab8a4eb421fa0d89540d86fceb1b7b7ce62edd3a578dd6d1709618ca0067"},"schema_version":"1.0"},"canonical_sha256":"fbf8c83e39b728a4ac20fae355c846e29e8ed32c83c65cb5d31723f179ba16fb","source":{"kind":"arxiv","id":"2605.11632","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.11632","created_at":"2026-06-05T01:14:40Z"},{"alias_kind":"arxiv_version","alias_value":"2605.11632v2","created_at":"2026-06-05T01:14:40Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.11632","created_at":"2026-06-05T01:14:40Z"},{"alias_kind":"pith_short_12","alias_value":"7P4MQPRZW4UK","created_at":"2026-06-05T01:14:40Z"},{"alias_kind":"pith_short_16","alias_value":"7P4MQPRZW4UKJLBA","created_at":"2026-06-05T01:14:40Z"},{"alias_kind":"pith_short_8","alias_value":"7P4MQPRZ","created_at":"2026-06-05T01:14:40Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:7P4MQPRZW4UKJLBA7LRVLSCG4K","target":"record","payload":{"canonical_record":{"source":{"id":"2605.11632","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-05-12T06:56:18Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"a6e1694b654f559b8075b950f5cc4ba25911443e16acdafd56ee72a1e31a47f8","abstract_canon_sha256":"6061ab8a4eb421fa0d89540d86fceb1b7b7ce62edd3a578dd6d1709618ca0067"},"schema_version":"1.0"},"canonical_sha256":"fbf8c83e39b728a4ac20fae355c846e29e8ed32c83c65cb5d31723f179ba16fb","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-05T01:14:40.563637Z","signature_b64":"sGBm3m+7ydb/XIqid27Rf38CRMAdRL1GMhK/xdmiKPwA2v8ywHiNwaWLncrZlGU1ormVZc20qB1Y6wnOyEMgCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"fbf8c83e39b728a4ac20fae355c846e29e8ed32c83c65cb5d31723f179ba16fb","last_reissued_at":"2026-06-05T01:14:40.562797Z","signature_status":"signed_v1","first_computed_at":"2026-06-05T01:14:40.562797Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2605.11632","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-06-05T01:14:40Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"qUk4ft2EopSNH+EczCmIP/XpzJwKbnS0NvdUjtNYW82fY4BSJYEyEdXtMyaGdRi5tiTdC6KQK6do1cwoiZy2Cw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-09T08:51:28.598355Z"},"content_sha256":"7898680e71181dee85c9ab86032344c2bb1864da58511db8d4fc002d34381e67","schema_version":"1.0","event_id":"sha256:7898680e71181dee85c9ab86032344c2bb1864da58511db8d4fc002d34381e67"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:7P4MQPRZW4UKJLBA7LRVLSCG4K","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Macro: Enhancing Multilingual Counterfactual Explanations through Alignment-as-Preference Optimization","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"A preference alignment method called Macro improves the validity of multilingual self-generated counterfactual explanations by 12.55 percent on average while maintaining minimality.","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Bohao Chu, Jing Yang, Qianli Wang, Simon Ostermann, Yihong Liu, Yilong Wang","submitted_at":"2026-05-12T06:56:18Z","abstract_excerpt":"Self-generated counterfactual explanations (SCEs) are minimally modified inputs (minimality) generated by large language models (LLMs) that flip their own predictions (validity), offering a causally grounded approach to unraveling black-box LLM behavior. Yet extending them beyond English remains challenging: existing methods struggle to produce valid SCEs in non-dominant languages, and a persistent trade-off between validity and minimality undermines explanation quality. We introduce Macro, a preference alignment framework that applies Direct Preference Optimization (DPO) to multilingual SCE g"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Experiments across four LLMs and seven typologically diverse languages show that Macro improves validity by 12.55% on average over the chain-of-thought baseline without degrading minimality, while avoiding the severe minimality violations of the translation-based baseline.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The composite scoring function used to construct preference pairs accurately and unbiasedly captures the validity-minimality trade-off across typologically diverse languages and different LLMs.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Macro uses Direct Preference Optimization on composite-scored preference pairs to improve validity of multilingual self-generated counterfactual explanations by 12.55% on average without degrading minimality.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"A preference alignment method called Macro improves the validity of multilingual self-generated counterfactual explanations by 12.55 percent on average while maintaining minimality.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"28e00cdb8af15cb3ccd9b8b0554f11a65c4b9b02c3c90dd8579829e53e2b0b6e"},"source":{"id":"2605.11632","kind":"arxiv","version":2},"verdict":{"id":"907fb810-b009-4583-a604-482d01beb33c","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-13T01:13:58.101109Z","strongest_claim":"Experiments across four LLMs and seven typologically diverse languages show that Macro improves validity by 12.55% on average over the chain-of-thought baseline without degrading minimality, while avoiding the severe minimality violations of the translation-based baseline.","one_line_summary":"Macro uses Direct Preference Optimization on composite-scored preference pairs to improve validity of multilingual self-generated counterfactual explanations by 12.55% on average without degrading minimality.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The composite scoring function used to construct preference pairs accurately and unbiasedly captures the validity-minimality trade-off across typologically diverse languages and different LLMs.","pith_extraction_headline":"A preference alignment method called Macro improves the validity of multilingual self-generated counterfactual explanations by 12.55 percent on average while maintaining minimality."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2605.11632/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"doi_title_agreement","ran_at":"2026-05-21T00:31:32.115026Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-20T14:16:48.159725Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"claim_evidence","ran_at":"2026-05-20T04:02:00.401849Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"ai_meta_artifact","ran_at":"2026-05-19T11:40:41.514511Z","status":"completed","version":"1.0.0","findings_count":0}],"snapshot_sha256":"c56bffa2b253209226b768852087a08432da90077ba9ddd798026b1ffe2beb6d"},"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":"907fb810-b009-4583-a604-482d01beb33c"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-06-05T01:14:40Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"2bCl0dV059Nmi9h3bVq8MfKDO01B63BT/4rpKD/9VaKZUmW0WpX5ttYqmn0YpF5ZSFXC81JGxw9yIb+ie65LDw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-09T08:51:28.599557Z"},"content_sha256":"2742721deef644ce0696f61a45897f213a284b7c10952ce51f3dd2998686a69f","schema_version":"1.0","event_id":"sha256:2742721deef644ce0696f61a45897f213a284b7c10952ce51f3dd2998686a69f"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/7P4MQPRZW4UKJLBA7LRVLSCG4K/bundle.json","state_url":"https://pith.science/pith/7P4MQPRZW4UKJLBA7LRVLSCG4K/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/7P4MQPRZW4UKJLBA7LRVLSCG4K/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":"2026-06-09T08:51:28Z","links":{"resolver":"https://pith.science/pith/7P4MQPRZW4UKJLBA7LRVLSCG4K","bundle":"https://pith.science/pith/7P4MQPRZW4UKJLBA7LRVLSCG4K/bundle.json","state":"https://pith.science/pith/7P4MQPRZW4UKJLBA7LRVLSCG4K/state.json","well_known_bundle":"https://pith.science/.well-known/pith/7P4MQPRZW4UKJLBA7LRVLSCG4K/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:7P4MQPRZW4UKJLBA7LRVLSCG4K","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":"6061ab8a4eb421fa0d89540d86fceb1b7b7ce62edd3a578dd6d1709618ca0067","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-05-12T06:56:18Z","title_canon_sha256":"a6e1694b654f559b8075b950f5cc4ba25911443e16acdafd56ee72a1e31a47f8"},"schema_version":"1.0","source":{"id":"2605.11632","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.11632","created_at":"2026-06-05T01:14:40Z"},{"alias_kind":"arxiv_version","alias_value":"2605.11632v2","created_at":"2026-06-05T01:14:40Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.11632","created_at":"2026-06-05T01:14:40Z"},{"alias_kind":"pith_short_12","alias_value":"7P4MQPRZW4UK","created_at":"2026-06-05T01:14:40Z"},{"alias_kind":"pith_short_16","alias_value":"7P4MQPRZW4UKJLBA","created_at":"2026-06-05T01:14:40Z"},{"alias_kind":"pith_short_8","alias_value":"7P4MQPRZ","created_at":"2026-06-05T01:14:40Z"}],"graph_snapshots":[{"event_id":"sha256:2742721deef644ce0696f61a45897f213a284b7c10952ce51f3dd2998686a69f","target":"graph","created_at":"2026-06-05T01:14:40Z","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":4,"items":[{"attestation":"unclaimed","claim_id":"C1","kind":"strongest_claim","source":"verdict.strongest_claim","status":"machine_extracted","text":"Experiments across four LLMs and seven typologically diverse languages show that Macro improves validity by 12.55% on average over the chain-of-thought baseline without degrading minimality, while avoiding the severe minimality violations of the translation-based baseline."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"The composite scoring function used to construct preference pairs accurately and unbiasedly captures the validity-minimality trade-off across typologically diverse languages and different LLMs."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"Macro uses Direct Preference Optimization on composite-scored preference pairs to improve validity of multilingual self-generated counterfactual explanations by 12.55% on average without degrading minimality."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"A preference alignment method called Macro improves the validity of multilingual self-generated counterfactual explanations by 12.55 percent on average while maintaining minimality."}],"snapshot_sha256":"28e00cdb8af15cb3ccd9b8b0554f11a65c4b9b02c3c90dd8579829e53e2b0b6e"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[{"findings_count":0,"name":"doi_title_agreement","ran_at":"2026-05-21T00:31:32.115026Z","status":"completed","version":"1.0.0"},{"findings_count":0,"name":"doi_compliance","ran_at":"2026-05-20T14:16:48.159725Z","status":"completed","version":"1.0.0"},{"findings_count":0,"name":"claim_evidence","ran_at":"2026-05-20T04:02:00.401849Z","status":"completed","version":"1.0.0"},{"findings_count":0,"name":"ai_meta_artifact","ran_at":"2026-05-19T11:40:41.514511Z","status":"completed","version":"1.0.0"}],"endpoint":"/pith/2605.11632/integrity.json","findings":[],"snapshot_sha256":"c56bffa2b253209226b768852087a08432da90077ba9ddd798026b1ffe2beb6d","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Self-generated counterfactual explanations (SCEs) are minimally modified inputs (minimality) generated by large language models (LLMs) that flip their own predictions (validity), offering a causally grounded approach to unraveling black-box LLM behavior. Yet extending them beyond English remains challenging: existing methods struggle to produce valid SCEs in non-dominant languages, and a persistent trade-off between validity and minimality undermines explanation quality. We introduce Macro, a preference alignment framework that applies Direct Preference Optimization (DPO) to multilingual SCE g","authors_text":"Bohao Chu, Jing Yang, Qianli Wang, Simon Ostermann, Yihong Liu, Yilong Wang","cross_cats":["cs.AI"],"headline":"A preference alignment method called Macro improves the validity of multilingual self-generated counterfactual explanations by 12.55 percent on average while maintaining minimality.","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-05-12T06:56:18Z","title":"Macro: Enhancing Multilingual Counterfactual Explanations through Alignment-as-Preference Optimization"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.11632","kind":"arxiv","version":2},"verdict":{"created_at":"2026-05-13T01:13:58.101109Z","id":"907fb810-b009-4583-a604-482d01beb33c","model_set":{"reader":"grok-4.3"},"one_line_summary":"Macro uses Direct Preference Optimization on composite-scored preference pairs to improve validity of multilingual self-generated counterfactual explanations by 12.55% on average without degrading minimality.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"A preference alignment method called Macro improves the validity of multilingual self-generated counterfactual explanations by 12.55 percent on average while maintaining minimality.","strongest_claim":"Experiments across four LLMs and seven typologically diverse languages show that Macro improves validity by 12.55% on average over the chain-of-thought baseline without degrading minimality, while avoiding the severe minimality violations of the translation-based baseline.","weakest_assumption":"The composite scoring function used to construct preference pairs accurately and unbiasedly captures the validity-minimality trade-off across typologically diverse languages and different LLMs."}},"verdict_id":"907fb810-b009-4583-a604-482d01beb33c"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:7898680e71181dee85c9ab86032344c2bb1864da58511db8d4fc002d34381e67","target":"record","created_at":"2026-06-05T01:14:40Z","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":"6061ab8a4eb421fa0d89540d86fceb1b7b7ce62edd3a578dd6d1709618ca0067","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-05-12T06:56:18Z","title_canon_sha256":"a6e1694b654f559b8075b950f5cc4ba25911443e16acdafd56ee72a1e31a47f8"},"schema_version":"1.0","source":{"id":"2605.11632","kind":"arxiv","version":2}},"canonical_sha256":"fbf8c83e39b728a4ac20fae355c846e29e8ed32c83c65cb5d31723f179ba16fb","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"fbf8c83e39b728a4ac20fae355c846e29e8ed32c83c65cb5d31723f179ba16fb","first_computed_at":"2026-06-05T01:14:40.562797Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-05T01:14:40.562797Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"sGBm3m+7ydb/XIqid27Rf38CRMAdRL1GMhK/xdmiKPwA2v8ywHiNwaWLncrZlGU1ormVZc20qB1Y6wnOyEMgCQ==","signature_status":"signed_v1","signed_at":"2026-06-05T01:14:40.563637Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.11632","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:7898680e71181dee85c9ab86032344c2bb1864da58511db8d4fc002d34381e67","sha256:2742721deef644ce0696f61a45897f213a284b7c10952ce51f3dd2998686a69f"],"state_sha256":"2740907393bf097d8b68317f801f1f8faa09abb96a1c1f01f8d4729e5b65ccf5"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"72tU+FYzXQgXaki31iJUqUzE0iKHEde5j+ufLmMC3uySkUS3skbelQHP4rVkrDBXajzPEBcp3hIxq2/DMSQJAA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-09T08:51:28.604106Z","bundle_sha256":"0d45b4e1a47b74fd04f38354e2bb2502f3d98ba278c03a9bf18811b26530e70d"}}