{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:YAECKVPUBUSLSD3LZNC3RI7PC7","short_pith_number":"pith:YAECKVPU","canonical_record":{"source":{"id":"2603.14405","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-03-15T14:38:32Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"ad37124e7a15daae1e9a235569ef151ede6cac444caf694fcb21ee0ccd25356a","abstract_canon_sha256":"6df421b6700acd7ac285085b20dba425b69126d190ed537fd2519dc21a5cfd83"},"schema_version":"1.0"},"canonical_sha256":"c0082555f40d24b90f6bcb45b8a3ef17c9ff8719cfda1ead9b497f1bb52aa169","source":{"kind":"arxiv","id":"2603.14405","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2603.14405","created_at":"2026-06-02T02:04:16Z"},{"alias_kind":"arxiv_version","alias_value":"2603.14405v2","created_at":"2026-06-02T02:04:16Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2603.14405","created_at":"2026-06-02T02:04:16Z"},{"alias_kind":"pith_short_12","alias_value":"YAECKVPUBUSL","created_at":"2026-06-02T02:04:16Z"},{"alias_kind":"pith_short_16","alias_value":"YAECKVPUBUSLSD3L","created_at":"2026-06-02T02:04:16Z"},{"alias_kind":"pith_short_8","alias_value":"YAECKVPU","created_at":"2026-06-02T02:04:16Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:YAECKVPUBUSLSD3LZNC3RI7PC7","target":"record","payload":{"canonical_record":{"source":{"id":"2603.14405","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-03-15T14:38:32Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"ad37124e7a15daae1e9a235569ef151ede6cac444caf694fcb21ee0ccd25356a","abstract_canon_sha256":"6df421b6700acd7ac285085b20dba425b69126d190ed537fd2519dc21a5cfd83"},"schema_version":"1.0"},"canonical_sha256":"c0082555f40d24b90f6bcb45b8a3ef17c9ff8719cfda1ead9b497f1bb52aa169","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-02T02:04:16.138645Z","signature_b64":"lBf9QwFkkQ5Y3mGX9Ifu8T+6PuGea+Z+RqGZobTmOo+N6eFbKUx7AvqBbC7lDzYEEyyPUml4TlI0lRsKi7V3DQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c0082555f40d24b90f6bcb45b8a3ef17c9ff8719cfda1ead9b497f1bb52aa169","last_reissued_at":"2026-06-02T02:04:16.138118Z","signature_status":"signed_v1","first_computed_at":"2026-06-02T02:04:16.138118Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2603.14405","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":"2026-06-02T02:04:16Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ZEprXB5pR2ZC/xFtM7lw9RsYNpW4Lyd2WfCFrfH1JRWAA+7bFNqSnG3F4pfgwTjySL0u7E+avvqNMjGPJjhzBA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-12T00:55:04.315010Z"},"content_sha256":"0cb3e32c75d07451f70f1e8518a348a76887a0de64439aaf16bcde1c6a7c9307","schema_version":"1.0","event_id":"sha256:0cb3e32c75d07451f70f1e8518a348a76887a0de64439aaf16bcde1c6a7c9307"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:YAECKVPUBUSLSD3LZNC3RI7PC7","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"ES-Merging: Biological MLLM Merging via Embedding Space Signals","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Dongki Kim, Sung Ju Hwang, Wonbin Lee","submitted_at":"2026-03-15T14:38:32Z","abstract_excerpt":"Biological multimodal large language models (MLLMs) have emerged as powerful foundation models for scientific discovery. However, existing models are specialized to a single modality, limiting their ability to solve inherently cross-modal scientific problems. While model merging is an efficient method to combine the different modalities into a unified MLLM, existing methods rely on input-agnostic parameter space heuristics that fail to faithfully capture modality specialization. To overcome this limitation, we propose the Embedding-Signal-based MLLM Merging (ES-Merging), a framework that estim"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2603.14405","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/2603.14405/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":"2026-06-02T02:04:16Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"++9Jgan4md/fkorRtG/d/iuCXErjWed48YTuwqcG1GYgMT1x8VgjKU85QPtiR9NcXOu1uGn+x97IHBpB+ur9CQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-12T00:55:04.315768Z"},"content_sha256":"65d5d6238d67fe5562c53909d94dda6c23c45d01e62b0af56c14ded9ad6a4037","schema_version":"1.0","event_id":"sha256:65d5d6238d67fe5562c53909d94dda6c23c45d01e62b0af56c14ded9ad6a4037"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/YAECKVPUBUSLSD3LZNC3RI7PC7/bundle.json","state_url":"https://pith.science/pith/YAECKVPUBUSLSD3LZNC3RI7PC7/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/YAECKVPUBUSLSD3LZNC3RI7PC7/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":"2026-06-12T00:55:04Z","links":{"resolver":"https://pith.science/pith/YAECKVPUBUSLSD3LZNC3RI7PC7","bundle":"https://pith.science/pith/YAECKVPUBUSLSD3LZNC3RI7PC7/bundle.json","state":"https://pith.science/pith/YAECKVPUBUSLSD3LZNC3RI7PC7/state.json","well_known_bundle":"https://pith.science/.well-known/pith/YAECKVPUBUSLSD3LZNC3RI7PC7/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:YAECKVPUBUSLSD3LZNC3RI7PC7","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":"6df421b6700acd7ac285085b20dba425b69126d190ed537fd2519dc21a5cfd83","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-03-15T14:38:32Z","title_canon_sha256":"ad37124e7a15daae1e9a235569ef151ede6cac444caf694fcb21ee0ccd25356a"},"schema_version":"1.0","source":{"id":"2603.14405","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2603.14405","created_at":"2026-06-02T02:04:16Z"},{"alias_kind":"arxiv_version","alias_value":"2603.14405v2","created_at":"2026-06-02T02:04:16Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2603.14405","created_at":"2026-06-02T02:04:16Z"},{"alias_kind":"pith_short_12","alias_value":"YAECKVPUBUSL","created_at":"2026-06-02T02:04:16Z"},{"alias_kind":"pith_short_16","alias_value":"YAECKVPUBUSLSD3L","created_at":"2026-06-02T02:04:16Z"},{"alias_kind":"pith_short_8","alias_value":"YAECKVPU","created_at":"2026-06-02T02:04:16Z"}],"graph_snapshots":[{"event_id":"sha256:65d5d6238d67fe5562c53909d94dda6c23c45d01e62b0af56c14ded9ad6a4037","target":"graph","created_at":"2026-06-02T02:04:16Z","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/2603.14405/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Biological multimodal large language models (MLLMs) have emerged as powerful foundation models for scientific discovery. However, existing models are specialized to a single modality, limiting their ability to solve inherently cross-modal scientific problems. While model merging is an efficient method to combine the different modalities into a unified MLLM, existing methods rely on input-agnostic parameter space heuristics that fail to faithfully capture modality specialization. To overcome this limitation, we propose the Embedding-Signal-based MLLM Merging (ES-Merging), a framework that estim","authors_text":"Dongki Kim, Sung Ju Hwang, Wonbin Lee","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-03-15T14:38:32Z","title":"ES-Merging: Biological MLLM Merging via Embedding Space Signals"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2603.14405","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":"sha256:0cb3e32c75d07451f70f1e8518a348a76887a0de64439aaf16bcde1c6a7c9307","target":"record","created_at":"2026-06-02T02:04:16Z","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":"6df421b6700acd7ac285085b20dba425b69126d190ed537fd2519dc21a5cfd83","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-03-15T14:38:32Z","title_canon_sha256":"ad37124e7a15daae1e9a235569ef151ede6cac444caf694fcb21ee0ccd25356a"},"schema_version":"1.0","source":{"id":"2603.14405","kind":"arxiv","version":2}},"canonical_sha256":"c0082555f40d24b90f6bcb45b8a3ef17c9ff8719cfda1ead9b497f1bb52aa169","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"c0082555f40d24b90f6bcb45b8a3ef17c9ff8719cfda1ead9b497f1bb52aa169","first_computed_at":"2026-06-02T02:04:16.138118Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-02T02:04:16.138118Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"lBf9QwFkkQ5Y3mGX9Ifu8T+6PuGea+Z+RqGZobTmOo+N6eFbKUx7AvqBbC7lDzYEEyyPUml4TlI0lRsKi7V3DQ==","signature_status":"signed_v1","signed_at":"2026-06-02T02:04:16.138645Z","signed_message":"canonical_sha256_bytes"},"source_id":"2603.14405","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:0cb3e32c75d07451f70f1e8518a348a76887a0de64439aaf16bcde1c6a7c9307","sha256:65d5d6238d67fe5562c53909d94dda6c23c45d01e62b0af56c14ded9ad6a4037"],"state_sha256":"a989a425ec2faa34a8b9840fac447a422617720dec0f434d95c19da4879222d7"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Ry0Q2ffldCFW+H+pTRb6sXnneIHe4IHn2nOh2qko0O1Ba2m/jqtERXcsDXd3lbHVboybiEFhFWGQjEprP2eeAA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-12T00:55:04.320217Z","bundle_sha256":"890170cc5d5aba64e054b2e6fbeef7820b5e0d6b75e022c9839bfd30e1af1ad9"}}