{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:P3IV4HAIV7R7QWKLKLRBN65H3J","short_pith_number":"pith:P3IV4HAI","schema_version":"1.0","canonical_sha256":"7ed15e1c08afe3f8594b52e216fba7da7b496a43f2623d920d3ffa3db4b7c8a7","source":{"kind":"arxiv","id":"2604.01674","version":2},"attestation_state":"computed","paper":{"title":"Can Heterogeneous Language Models Be Fused?","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Jie Zhou, Liang He, Qi Feng, Qin Chen, Shilian Chen, Wen Wu, Xin Li","submitted_at":"2026-04-02T06:21:21Z","abstract_excerpt":"Model merging aims to integrate multiple expert models into a single model that inherits their complementary strengths without incurring the inference-time cost of ensembling. Recent progress has shown that merging can be highly effective when all source models are \\emph{homogeneous}, i.e., derived from the same pretrained backbone and therefore share aligned parameter coordinates or compatible task vectors. Yet this assumption is increasingly unrealistic in open model ecosystems, where useful experts are often built on different families such as Llama, Qwen, and Mistral. In such \\emph{heterog"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2604.01674","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-04-02T06:21:21Z","cross_cats_sorted":[],"title_canon_sha256":"bffd0fbebbf0b3131563d4e45c7dac3d1e17fb0150e3fe81d49f3ac458b5d776","abstract_canon_sha256":"851929f28e6cf324126e1b0118035caa6561eb8c39401d9fe0a8cce77ba9a6f7"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T00:03:10.330929Z","signature_b64":"SSFVAjkcbYsdbPvUwzSbxVrYiA0NE3JAIDlREXxjBKDf0yThIDQssixBf+2FMsYhqYpSff+d0JtBozGSU/N3Bw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7ed15e1c08afe3f8594b52e216fba7da7b496a43f2623d920d3ffa3db4b7c8a7","last_reissued_at":"2026-05-20T00:03:10.329988Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T00:03:10.329988Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Can Heterogeneous Language Models Be Fused?","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Jie Zhou, Liang He, Qi Feng, Qin Chen, Shilian Chen, Wen Wu, Xin Li","submitted_at":"2026-04-02T06:21:21Z","abstract_excerpt":"Model merging aims to integrate multiple expert models into a single model that inherits their complementary strengths without incurring the inference-time cost of ensembling. Recent progress has shown that merging can be highly effective when all source models are \\emph{homogeneous}, i.e., derived from the same pretrained backbone and therefore share aligned parameter coordinates or compatible task vectors. Yet this assumption is increasingly unrealistic in open model ecosystems, where useful experts are often built on different families such as Llama, Qwen, and Mistral. In such \\emph{heterog"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2604.01674","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/2604.01674/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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2604.01674","created_at":"2026-05-20T00:03:10.330114+00:00"},{"alias_kind":"arxiv_version","alias_value":"2604.01674v2","created_at":"2026-05-20T00:03:10.330114+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2604.01674","created_at":"2026-05-20T00:03:10.330114+00:00"},{"alias_kind":"pith_short_12","alias_value":"P3IV4HAIV7R7","created_at":"2026-05-20T00:03:10.330114+00:00"},{"alias_kind":"pith_short_16","alias_value":"P3IV4HAIV7R7QWKL","created_at":"2026-05-20T00:03:10.330114+00:00"},{"alias_kind":"pith_short_8","alias_value":"P3IV4HAI","created_at":"2026-05-20T00:03:10.330114+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/P3IV4HAIV7R7QWKLKLRBN65H3J","json":"https://pith.science/pith/P3IV4HAIV7R7QWKLKLRBN65H3J.json","graph_json":"https://pith.science/api/pith-number/P3IV4HAIV7R7QWKLKLRBN65H3J/graph.json","events_json":"https://pith.science/api/pith-number/P3IV4HAIV7R7QWKLKLRBN65H3J/events.json","paper":"https://pith.science/paper/P3IV4HAI"},"agent_actions":{"view_html":"https://pith.science/pith/P3IV4HAIV7R7QWKLKLRBN65H3J","download_json":"https://pith.science/pith/P3IV4HAIV7R7QWKLKLRBN65H3J.json","view_paper":"https://pith.science/paper/P3IV4HAI","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2604.01674&json=true","fetch_graph":"https://pith.science/api/pith-number/P3IV4HAIV7R7QWKLKLRBN65H3J/graph.json","fetch_events":"https://pith.science/api/pith-number/P3IV4HAIV7R7QWKLKLRBN65H3J/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/P3IV4HAIV7R7QWKLKLRBN65H3J/action/timestamp_anchor","attest_storage":"https://pith.science/pith/P3IV4HAIV7R7QWKLKLRBN65H3J/action/storage_attestation","attest_author":"https://pith.science/pith/P3IV4HAIV7R7QWKLKLRBN65H3J/action/author_attestation","sign_citation":"https://pith.science/pith/P3IV4HAIV7R7QWKLKLRBN65H3J/action/citation_signature","submit_replication":"https://pith.science/pith/P3IV4HAIV7R7QWKLKLRBN65H3J/action/replication_record"}},"created_at":"2026-05-20T00:03:10.330114+00:00","updated_at":"2026-05-20T00:03:10.330114+00:00"}