{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:Y6EIBTXUSHE63647ESX2VGGIY2","short_pith_number":"pith:Y6EIBTXU","canonical_record":{"source":{"id":"2512.16300","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2025-12-18T08:38:44Z","cross_cats_sorted":[],"title_canon_sha256":"41a8b80090e0e708c2bc4920925f5998d4c1bd534e2cb830fb5924225a39246e","abstract_canon_sha256":"a903e3e6e682262cbf0ec94d2c36004c7aae353a08772f75d08e68cfd5b4e520"},"schema_version":"1.0"},"canonical_sha256":"c78880cef491c9edfb9f24afaa98c8c6b59c6cd80f639f4f85b9311b186a04f5","source":{"kind":"arxiv","id":"2512.16300","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2512.16300","created_at":"2026-07-23T01:24:19Z"},{"alias_kind":"arxiv_version","alias_value":"2512.16300v3","created_at":"2026-07-23T01:24:19Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2512.16300","created_at":"2026-07-23T01:24:19Z"},{"alias_kind":"pith_short_12","alias_value":"Y6EIBTXUSHE6","created_at":"2026-07-23T01:24:19Z"},{"alias_kind":"pith_short_16","alias_value":"Y6EIBTXUSHE63647","created_at":"2026-07-23T01:24:19Z"},{"alias_kind":"pith_short_8","alias_value":"Y6EIBTXU","created_at":"2026-07-23T01:24:19Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:Y6EIBTXUSHE63647ESX2VGGIY2","target":"record","payload":{"canonical_record":{"source":{"id":"2512.16300","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2025-12-18T08:38:44Z","cross_cats_sorted":[],"title_canon_sha256":"41a8b80090e0e708c2bc4920925f5998d4c1bd534e2cb830fb5924225a39246e","abstract_canon_sha256":"a903e3e6e682262cbf0ec94d2c36004c7aae353a08772f75d08e68cfd5b4e520"},"schema_version":"1.0"},"canonical_sha256":"c78880cef491c9edfb9f24afaa98c8c6b59c6cd80f639f4f85b9311b186a04f5","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-23T01:24:19.610987Z","signature_b64":"60ItJvTdPGrevUODwWUgCfJroMcmSwmMXTHRx2O3i7594IBpyQKtvNmzxDv/c1Nhnbven6uPjciLfXlOrT9WDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c78880cef491c9edfb9f24afaa98c8c6b59c6cd80f639f4f85b9311b186a04f5","last_reissued_at":"2026-07-23T01:24:19.610072Z","signature_status":"signed_v1","first_computed_at":"2026-07-23T01:24:19.610072Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2512.16300","source_version":3,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-23T01:24:19Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"A0EhIY1A4/Jt8ULCNHKcD+G0argyGTSD8Ikdx9brt4dGPUd00e5vfCGV7EBaBkNN1a853wOEXebyRvB3YAzZCA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-05T03:03:47.269233Z"},"content_sha256":"e16e94c9aafe5ea4bdab031f01c371a3383be3d86c68ae231182b2e34d3317a7","schema_version":"1.0","event_id":"sha256:e16e94c9aafe5ea4bdab031f01c371a3383be3d86c68ae231182b2e34d3317a7"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:Y6EIBTXUSHE63647ESX2VGGIY2","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Code-in-the-Loop Forensics: Agentic Tool Use for Image Forgery Detection","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"Multimodal language models can generate and run their own Python tools to detect image forgeries by combining low-level artifacts with semantic reasoning.","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Chuanhao Li, Fanrui Zhang, Jianwen Sun, Jiawei Liu, Jiaxin Ai, Kaipeng Zhang, Qiang Zhang, Sizhuo Zhou, Wenjie Li, Yifan Chang, Yujie Zhang, Yukang Feng, Zizhen Li","submitted_at":"2025-12-18T08:38:44Z","abstract_excerpt":"Existing image forgery detection (IFD) methods either exploit low-level, semantics-agnostic artifacts or rely on multimodal large language models (MLLMs) with high-level semantic knowledge. Although naturally complementary, these two information streams are highly heterogeneous in both paradigm and reasoning, making it difficult for existing methods to unify them or effectively model their cross-level interactions. To address this gap, we propose ForenAgent, a multi-round interactive IFD framework that enables MLLMs to autonomously generate, execute, and iteratively refine Python-based low-lev"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"ForenAgent exhibits emergent tool-use competence and reflective reasoning on challenging IFD tasks when assisted by low-level tools, charting a promising route toward general-purpose IFD.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That multimodal LLMs can reliably generate correct, executable, and useful Python code for low-level image analysis tasks without introducing errors or hallucinations that undermine the detection process.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"ForenAgent lets MLLMs create and iteratively improve low-level Python tools for image forgery detection via a two-stage training pipeline and a new 100k-image benchmark dataset.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Multimodal language models can generate and run their own Python tools to detect image forgeries by combining low-level artifacts with semantic reasoning.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"8b6d8e0532fc6d5a8cac75e83164e9b9a67f5ef786038f3948f75af95b5bf8c7"},"source":{"id":"2512.16300","kind":"arxiv","version":3},"verdict":{"id":"62886a62-fc6f-4518-9f51-8fd3ec3f4f0d","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-16T21:37:39.471784Z","strongest_claim":"ForenAgent exhibits emergent tool-use competence and reflective reasoning on challenging IFD tasks when assisted by low-level tools, charting a promising route toward general-purpose IFD.","one_line_summary":"ForenAgent lets MLLMs create and iteratively improve low-level Python tools for image forgery detection via a two-stage training pipeline and a new 100k-image benchmark dataset.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That multimodal LLMs can reliably generate correct, executable, and useful Python code for low-level image analysis tasks without introducing errors or hallucinations that undermine the detection process.","pith_extraction_headline":"Multimodal language models can generate and run their own Python tools to detect image forgeries by combining low-level artifacts with semantic reasoning."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2512.16300/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":2,"snapshot_sha256":"1ade4a986e3c345cbf6ff9e2ffeadb52af4f33afdc3694c9337f03c171554998"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"verdict_id":"62886a62-fc6f-4518-9f51-8fd3ec3f4f0d"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-23T01:24:19Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"yhhY/mjcatdhmiIweiAfjDH4hm1uD4CThMf4rND1UWLu1s3jVJD15UixG6ujcuLyYykgndeuLFgCxbP3k1MXDQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-05T03:03:47.270431Z"},"content_sha256":"8dabc650f59d1d1b3180371e2da5af7ba5e9f768cd04e1a37703a8ed42e59f3e","schema_version":"1.0","event_id":"sha256:8dabc650f59d1d1b3180371e2da5af7ba5e9f768cd04e1a37703a8ed42e59f3e"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/Y6EIBTXUSHE63647ESX2VGGIY2/bundle.json","state_url":"https://pith.science/pith/Y6EIBTXUSHE63647ESX2VGGIY2/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/Y6EIBTXUSHE63647ESX2VGGIY2/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":"2026-08-05T03:03:47Z","links":{"resolver":"https://pith.science/pith/Y6EIBTXUSHE63647ESX2VGGIY2","bundle":"https://pith.science/pith/Y6EIBTXUSHE63647ESX2VGGIY2/bundle.json","state":"https://pith.science/pith/Y6EIBTXUSHE63647ESX2VGGIY2/state.json","well_known_bundle":"https://pith.science/.well-known/pith/Y6EIBTXUSHE63647ESX2VGGIY2/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:Y6EIBTXUSHE63647ESX2VGGIY2","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":"a903e3e6e682262cbf0ec94d2c36004c7aae353a08772f75d08e68cfd5b4e520","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2025-12-18T08:38:44Z","title_canon_sha256":"41a8b80090e0e708c2bc4920925f5998d4c1bd534e2cb830fb5924225a39246e"},"schema_version":"1.0","source":{"id":"2512.16300","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2512.16300","created_at":"2026-07-23T01:24:19Z"},{"alias_kind":"arxiv_version","alias_value":"2512.16300v3","created_at":"2026-07-23T01:24:19Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2512.16300","created_at":"2026-07-23T01:24:19Z"},{"alias_kind":"pith_short_12","alias_value":"Y6EIBTXUSHE6","created_at":"2026-07-23T01:24:19Z"},{"alias_kind":"pith_short_16","alias_value":"Y6EIBTXUSHE63647","created_at":"2026-07-23T01:24:19Z"},{"alias_kind":"pith_short_8","alias_value":"Y6EIBTXU","created_at":"2026-07-23T01:24:19Z"}],"graph_snapshots":[{"event_id":"sha256:8dabc650f59d1d1b3180371e2da5af7ba5e9f768cd04e1a37703a8ed42e59f3e","target":"graph","created_at":"2026-07-23T01:24:19Z","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":"ForenAgent exhibits emergent tool-use competence and reflective reasoning on challenging IFD tasks when assisted by low-level tools, charting a promising route toward general-purpose IFD."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"That multimodal LLMs can reliably generate correct, executable, and useful Python code for low-level image analysis tasks without introducing errors or hallucinations that undermine the detection process."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"ForenAgent lets MLLMs create and iteratively improve low-level Python tools for image forgery detection via a two-stage training pipeline and a new 100k-image benchmark dataset."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"Multimodal language models can generate and run their own Python tools to detect image forgeries by combining low-level artifacts with semantic reasoning."}],"snapshot_sha256":"8b6d8e0532fc6d5a8cac75e83164e9b9a67f5ef786038f3948f75af95b5bf8c7"},"formal_canon":{"evidence_count":2,"snapshot_sha256":"1ade4a986e3c345cbf6ff9e2ffeadb52af4f33afdc3694c9337f03c171554998"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2512.16300/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Existing image forgery detection (IFD) methods either exploit low-level, semantics-agnostic artifacts or rely on multimodal large language models (MLLMs) with high-level semantic knowledge. Although naturally complementary, these two information streams are highly heterogeneous in both paradigm and reasoning, making it difficult for existing methods to unify them or effectively model their cross-level interactions. To address this gap, we propose ForenAgent, a multi-round interactive IFD framework that enables MLLMs to autonomously generate, execute, and iteratively refine Python-based low-lev","authors_text":"Chuanhao Li, Fanrui Zhang, Jianwen Sun, Jiawei Liu, Jiaxin Ai, Kaipeng Zhang, Qiang Zhang, Sizhuo Zhou, Wenjie Li, Yifan Chang, Yujie Zhang, Yukang Feng, Zizhen Li","cross_cats":[],"headline":"Multimodal language models can generate and run their own Python tools to detect image forgeries by combining low-level artifacts with semantic reasoning.","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2025-12-18T08:38:44Z","title":"Code-in-the-Loop Forensics: Agentic Tool Use for Image Forgery Detection"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2512.16300","kind":"arxiv","version":3},"verdict":{"created_at":"2026-05-16T21:37:39.471784Z","id":"62886a62-fc6f-4518-9f51-8fd3ec3f4f0d","model_set":{"reader":"grok-4.3"},"one_line_summary":"ForenAgent lets MLLMs create and iteratively improve low-level Python tools for image forgery detection via a two-stage training pipeline and a new 100k-image benchmark dataset.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"Multimodal language models can generate and run their own Python tools to detect image forgeries by combining low-level artifacts with semantic reasoning.","strongest_claim":"ForenAgent exhibits emergent tool-use competence and reflective reasoning on challenging IFD tasks when assisted by low-level tools, charting a promising route toward general-purpose IFD.","weakest_assumption":"That multimodal LLMs can reliably generate correct, executable, and useful Python code for low-level image analysis tasks without introducing errors or hallucinations that undermine the detection process."}},"verdict_id":"62886a62-fc6f-4518-9f51-8fd3ec3f4f0d"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:e16e94c9aafe5ea4bdab031f01c371a3383be3d86c68ae231182b2e34d3317a7","target":"record","created_at":"2026-07-23T01:24:19Z","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":"a903e3e6e682262cbf0ec94d2c36004c7aae353a08772f75d08e68cfd5b4e520","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2025-12-18T08:38:44Z","title_canon_sha256":"41a8b80090e0e708c2bc4920925f5998d4c1bd534e2cb830fb5924225a39246e"},"schema_version":"1.0","source":{"id":"2512.16300","kind":"arxiv","version":3}},"canonical_sha256":"c78880cef491c9edfb9f24afaa98c8c6b59c6cd80f639f4f85b9311b186a04f5","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"c78880cef491c9edfb9f24afaa98c8c6b59c6cd80f639f4f85b9311b186a04f5","first_computed_at":"2026-07-23T01:24:19.610072Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-23T01:24:19.610072Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"60ItJvTdPGrevUODwWUgCfJroMcmSwmMXTHRx2O3i7594IBpyQKtvNmzxDv/c1Nhnbven6uPjciLfXlOrT9WDg==","signature_status":"signed_v1","signed_at":"2026-07-23T01:24:19.610987Z","signed_message":"canonical_sha256_bytes"},"source_id":"2512.16300","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:e16e94c9aafe5ea4bdab031f01c371a3383be3d86c68ae231182b2e34d3317a7","sha256:8dabc650f59d1d1b3180371e2da5af7ba5e9f768cd04e1a37703a8ed42e59f3e"],"state_sha256":"34512c00b9dee70b7683560c7b056d3817533311762d788ef4fc6a217327dda7"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"eV71ZrH4vSOdBo3h6t0fn9pA3ypA6KPqDQzV6wZEdRPpxJsMAhYSPR1Z8Q6HZsm9HqA9C3ARFYQvfnlN7Yk8AQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-05T03:03:47.277668Z","bundle_sha256":"2d2ba8f7647d504bb3a7345891985423d702e69ad0568a03752fb74a8da10972"}}