{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:CXS4ADK75S3QIE2AHUSJGFTRJQ","short_pith_number":"pith:CXS4ADK7","canonical_record":{"source":{"id":"2509.01535","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-09-01T15:13:15Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"04578fc105a2fbb2e42f0c29cd83126d8797fc295fd198a97dc8c6b0e2de8776","abstract_canon_sha256":"a18c2ff7cfbe935c171b91f9ca0e50f0be8ba37540ae6ff6e8b906a61bf142c1"},"schema_version":"1.0"},"canonical_sha256":"15e5c00d5fecb70413403d249316714c14936d3fa1ca1f8faadc18905b9258c7","source":{"kind":"arxiv","id":"2509.01535","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2509.01535","created_at":"2026-07-05T12:07:11Z"},{"alias_kind":"arxiv_version","alias_value":"2509.01535v2","created_at":"2026-07-05T12:07:11Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2509.01535","created_at":"2026-07-05T12:07:11Z"},{"alias_kind":"pith_short_12","alias_value":"CXS4ADK75S3Q","created_at":"2026-07-05T12:07:11Z"},{"alias_kind":"pith_short_16","alias_value":"CXS4ADK75S3QIE2A","created_at":"2026-07-05T12:07:11Z"},{"alias_kind":"pith_short_8","alias_value":"CXS4ADK7","created_at":"2026-07-05T12:07:11Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:CXS4ADK75S3QIE2AHUSJGFTRJQ","target":"record","payload":{"canonical_record":{"source":{"id":"2509.01535","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-09-01T15:13:15Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"04578fc105a2fbb2e42f0c29cd83126d8797fc295fd198a97dc8c6b0e2de8776","abstract_canon_sha256":"a18c2ff7cfbe935c171b91f9ca0e50f0be8ba37540ae6ff6e8b906a61bf142c1"},"schema_version":"1.0"},"canonical_sha256":"15e5c00d5fecb70413403d249316714c14936d3fa1ca1f8faadc18905b9258c7","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T12:07:11.418802Z","signature_b64":"k9dnwmDP+1p32v4zfOySfOpP1mTVsPazd/Ea43CSCB365idJ2TvuMdx2NpwDOR2HPuSx/kqb+ladwfYVunRfBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"15e5c00d5fecb70413403d249316714c14936d3fa1ca1f8faadc18905b9258c7","last_reissued_at":"2026-07-05T12:07:11.418374Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T12:07:11.418374Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2509.01535","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":"2026-07-05T12:07:11Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"cc/b33bvQVQ86caMhLTfjjlL4e7siiNVIjzO55p1hrYdGf8CnDO8TZw0yTIYC5CIN0fr53FnqFIfX3CnyyB/BA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-06T03:02:18.068328Z"},"content_sha256":"6a6662f7225118f2c1b807a525ec163197e97e271a3e874d0e278adea8676786","schema_version":"1.0","event_id":"sha256:6a6662f7225118f2c1b807a525ec163197e97e271a3e874d0e278adea8676786"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:CXS4ADK75S3QIE2AHUSJGFTRJQ","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"CAT: Causal Attention Tuning For Injecting Fine-grained Causal Knowledge into Large Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Junjian Ye, Kairong Han, Kun Kuang, Lujia Pan, Wenshuo Zhao, Ziyu Zhao","submitted_at":"2025-09-01T15:13:15Z","abstract_excerpt":"Large Language Models (LLMs) have achieved remarkable success across various domains. However, a fundamental question remains: Can LLMs effectively utilize causal knowledge for prediction and generation? Through empirical studies, we find that LLMs trained directly on large-scale data often capture spurious correlations rather than true causal relationships, leading to suboptimal performance, especially in out-of-distribution (OOD) scenarios. To address this challenge, we propose Causal Attention Tuning (CAT), a novel approach that injects fine-grained causal knowledge into the attention mecha"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2509.01535","kind":"arxiv","version":2},"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/2509.01535/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":"2026-07-05T12:07:11Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"TisyOzuKuB1Srfltng0uJaUFiLwlJJlZET3wMH1dIZbdKVRU63HXhjSVqCsFsESJ1G0LkrSwH9b9YekgP/WDDw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-06T03:02:18.068961Z"},"content_sha256":"ca441c5ef61902b288d50e068b45f39ac73f944e36831d4e069a58bf8f456d06","schema_version":"1.0","event_id":"sha256:ca441c5ef61902b288d50e068b45f39ac73f944e36831d4e069a58bf8f456d06"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/CXS4ADK75S3QIE2AHUSJGFTRJQ/bundle.json","state_url":"https://pith.science/pith/CXS4ADK75S3QIE2AHUSJGFTRJQ/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/CXS4ADK75S3QIE2AHUSJGFTRJQ/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":"2026-08-06T03:02:18Z","links":{"resolver":"https://pith.science/pith/CXS4ADK75S3QIE2AHUSJGFTRJQ","bundle":"https://pith.science/pith/CXS4ADK75S3QIE2AHUSJGFTRJQ/bundle.json","state":"https://pith.science/pith/CXS4ADK75S3QIE2AHUSJGFTRJQ/state.json","well_known_bundle":"https://pith.science/.well-known/pith/CXS4ADK75S3QIE2AHUSJGFTRJQ/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:CXS4ADK75S3QIE2AHUSJGFTRJQ","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":"a18c2ff7cfbe935c171b91f9ca0e50f0be8ba37540ae6ff6e8b906a61bf142c1","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-09-01T15:13:15Z","title_canon_sha256":"04578fc105a2fbb2e42f0c29cd83126d8797fc295fd198a97dc8c6b0e2de8776"},"schema_version":"1.0","source":{"id":"2509.01535","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2509.01535","created_at":"2026-07-05T12:07:11Z"},{"alias_kind":"arxiv_version","alias_value":"2509.01535v2","created_at":"2026-07-05T12:07:11Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2509.01535","created_at":"2026-07-05T12:07:11Z"},{"alias_kind":"pith_short_12","alias_value":"CXS4ADK75S3Q","created_at":"2026-07-05T12:07:11Z"},{"alias_kind":"pith_short_16","alias_value":"CXS4ADK75S3QIE2A","created_at":"2026-07-05T12:07:11Z"},{"alias_kind":"pith_short_8","alias_value":"CXS4ADK7","created_at":"2026-07-05T12:07:11Z"}],"graph_snapshots":[{"event_id":"sha256:ca441c5ef61902b288d50e068b45f39ac73f944e36831d4e069a58bf8f456d06","target":"graph","created_at":"2026-07-05T12:07:11Z","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/2509.01535/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large Language Models (LLMs) have achieved remarkable success across various domains. However, a fundamental question remains: Can LLMs effectively utilize causal knowledge for prediction and generation? Through empirical studies, we find that LLMs trained directly on large-scale data often capture spurious correlations rather than true causal relationships, leading to suboptimal performance, especially in out-of-distribution (OOD) scenarios. To address this challenge, we propose Causal Attention Tuning (CAT), a novel approach that injects fine-grained causal knowledge into the attention mecha","authors_text":"Junjian Ye, Kairong Han, Kun Kuang, Lujia Pan, Wenshuo Zhao, Ziyu Zhao","cross_cats":["cs.AI"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-09-01T15:13:15Z","title":"CAT: Causal Attention Tuning For Injecting Fine-grained Causal Knowledge into Large Language Models"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2509.01535","kind":"arxiv","version":2},"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":"sha256:6a6662f7225118f2c1b807a525ec163197e97e271a3e874d0e278adea8676786","target":"record","created_at":"2026-07-05T12:07:11Z","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":"a18c2ff7cfbe935c171b91f9ca0e50f0be8ba37540ae6ff6e8b906a61bf142c1","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-09-01T15:13:15Z","title_canon_sha256":"04578fc105a2fbb2e42f0c29cd83126d8797fc295fd198a97dc8c6b0e2de8776"},"schema_version":"1.0","source":{"id":"2509.01535","kind":"arxiv","version":2}},"canonical_sha256":"15e5c00d5fecb70413403d249316714c14936d3fa1ca1f8faadc18905b9258c7","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"15e5c00d5fecb70413403d249316714c14936d3fa1ca1f8faadc18905b9258c7","first_computed_at":"2026-07-05T12:07:11.418374Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T12:07:11.418374Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"k9dnwmDP+1p32v4zfOySfOpP1mTVsPazd/Ea43CSCB365idJ2TvuMdx2NpwDOR2HPuSx/kqb+ladwfYVunRfBw==","signature_status":"signed_v1","signed_at":"2026-07-05T12:07:11.418802Z","signed_message":"canonical_sha256_bytes"},"source_id":"2509.01535","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:6a6662f7225118f2c1b807a525ec163197e97e271a3e874d0e278adea8676786","sha256:ca441c5ef61902b288d50e068b45f39ac73f944e36831d4e069a58bf8f456d06"],"state_sha256":"023c1d0ccd463e184592f334d88a33de69108bd9658f8a147f4e9fbc820ed3f2"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"CgwFiMJEBiGSl+8/R19EA3p8BhVxM2qj+ytgmU0ZcaQyozbN+Vsm/9JScX3j7S0lMoPiaqThzuCAx6984GAcDA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-06T03:02:18.075184Z","bundle_sha256":"c0143b0932abb57306f040cfcf5caa25b86ac0ff652c1ca42421d90b81e1139b"}}