{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:IOAIZMQW5QYG5DON64ZO5EMRP6","short_pith_number":"pith:IOAIZMQW","canonical_record":{"source":{"id":"2606.17949","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.DC","submitted_at":"2026-06-16T14:02:34Z","cross_cats_sorted":[],"title_canon_sha256":"bf1af85a692642e400b21f38fa200e92675da516f80ced4fb07756698b57cc78","abstract_canon_sha256":"8275f93eb0c8062d0b50e1c351de1bb1d63517a27fcac44d38a58f2b14da8c13"},"schema_version":"1.0"},"canonical_sha256":"43808cb216ec306e8dcdf732ee91917fb565a8a7d1b92aaa189a5b8eb4125f27","source":{"kind":"arxiv","id":"2606.17949","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.17949","created_at":"2026-06-19T16:10:44Z"},{"alias_kind":"arxiv_version","alias_value":"2606.17949v1","created_at":"2026-06-19T16:10:44Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.17949","created_at":"2026-06-19T16:10:44Z"},{"alias_kind":"pith_short_12","alias_value":"IOAIZMQW5QYG","created_at":"2026-06-19T16:10:44Z"},{"alias_kind":"pith_short_16","alias_value":"IOAIZMQW5QYG5DON","created_at":"2026-06-19T16:10:44Z"},{"alias_kind":"pith_short_8","alias_value":"IOAIZMQW","created_at":"2026-06-19T16:10:44Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:IOAIZMQW5QYG5DON64ZO5EMRP6","target":"record","payload":{"canonical_record":{"source":{"id":"2606.17949","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.DC","submitted_at":"2026-06-16T14:02:34Z","cross_cats_sorted":[],"title_canon_sha256":"bf1af85a692642e400b21f38fa200e92675da516f80ced4fb07756698b57cc78","abstract_canon_sha256":"8275f93eb0c8062d0b50e1c351de1bb1d63517a27fcac44d38a58f2b14da8c13"},"schema_version":"1.0"},"canonical_sha256":"43808cb216ec306e8dcdf732ee91917fb565a8a7d1b92aaa189a5b8eb4125f27","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-19T16:10:44.310431Z","signature_b64":"Zl8xXlhB9ziAO61EkUyaBFQPQXbN3kANWeo7YFQWOjhRbr43/3dzAGWc9cVpoN/xp6UiSZV5y972z6LNmSSRCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"43808cb216ec306e8dcdf732ee91917fb565a8a7d1b92aaa189a5b8eb4125f27","last_reissued_at":"2026-06-19T16:10:44.310069Z","signature_status":"signed_v1","first_computed_at":"2026-06-19T16:10:44.310069Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2606.17949","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-06-19T16:10:44Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"aVzxPVusycV8Zq+EnNU5uIqRZDtdlnLsggrZU9DZbPKOr0YGpdP9zMObmP0Ht0k2dgetSc+dBB3tD54TQ5mwCg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-04T20:50:39.117900Z"},"content_sha256":"b687ee8424c58299650c3de38df15c3ffbe5e7f724bf2b457d2c88f7162d501b","schema_version":"1.0","event_id":"sha256:b687ee8424c58299650c3de38df15c3ffbe5e7f724bf2b457d2c88f7162d501b"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:IOAIZMQW5QYG5DON64ZO5EMRP6","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"RouteBalance: Fused Model Routing and Load Balancing for Heterogeneous LLM Serving","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.DC","authors_text":"Evangelia Kalyvianaki, Wei Da","submitted_at":"2026-06-16T14:02:34Z","abstract_excerpt":"Heterogeneous LLM serving stacks split scheduling into two layers that optimize in isolation: model routers pick a model from quality and cost signals while ignoring instance load, and serving load balancers optimize queues while ignoring quality. We present RouteBalance, a serving-aware scheduling layer that fuses both into a single online assignment over concrete model instances, jointly trading off quality, latency, and cost. A batched in-process predictor stack and dead-reckoned instance state keep the joint decision cheap on the request hot path ($\\approx$32 ms at 12 req/s). On a 13-insta"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.17949","kind":"arxiv","version":1},"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/2606.17949/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":"2026-06-19T16:10:44Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"h9gI6aWVAqqA3CUtLxY90lu0X+KENpwiQnG/7zzJNFcwbCSEsXZxnAAu5Y8RvdRm/O1QuXYnJdxD9r8GAbxGBQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-04T20:50:39.118262Z"},"content_sha256":"008d45d65792600dec285b3839676971275e2b4a8ae32d9b82eb80215c7489fc","schema_version":"1.0","event_id":"sha256:008d45d65792600dec285b3839676971275e2b4a8ae32d9b82eb80215c7489fc"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/IOAIZMQW5QYG5DON64ZO5EMRP6/bundle.json","state_url":"https://pith.science/pith/IOAIZMQW5QYG5DON64ZO5EMRP6/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/IOAIZMQW5QYG5DON64ZO5EMRP6/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":"2026-07-04T20:50:39Z","links":{"resolver":"https://pith.science/pith/IOAIZMQW5QYG5DON64ZO5EMRP6","bundle":"https://pith.science/pith/IOAIZMQW5QYG5DON64ZO5EMRP6/bundle.json","state":"https://pith.science/pith/IOAIZMQW5QYG5DON64ZO5EMRP6/state.json","well_known_bundle":"https://pith.science/.well-known/pith/IOAIZMQW5QYG5DON64ZO5EMRP6/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:IOAIZMQW5QYG5DON64ZO5EMRP6","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":"8275f93eb0c8062d0b50e1c351de1bb1d63517a27fcac44d38a58f2b14da8c13","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.DC","submitted_at":"2026-06-16T14:02:34Z","title_canon_sha256":"bf1af85a692642e400b21f38fa200e92675da516f80ced4fb07756698b57cc78"},"schema_version":"1.0","source":{"id":"2606.17949","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.17949","created_at":"2026-06-19T16:10:44Z"},{"alias_kind":"arxiv_version","alias_value":"2606.17949v1","created_at":"2026-06-19T16:10:44Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.17949","created_at":"2026-06-19T16:10:44Z"},{"alias_kind":"pith_short_12","alias_value":"IOAIZMQW5QYG","created_at":"2026-06-19T16:10:44Z"},{"alias_kind":"pith_short_16","alias_value":"IOAIZMQW5QYG5DON","created_at":"2026-06-19T16:10:44Z"},{"alias_kind":"pith_short_8","alias_value":"IOAIZMQW","created_at":"2026-06-19T16:10:44Z"}],"graph_snapshots":[{"event_id":"sha256:008d45d65792600dec285b3839676971275e2b4a8ae32d9b82eb80215c7489fc","target":"graph","created_at":"2026-06-19T16:10:44Z","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/2606.17949/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Heterogeneous LLM serving stacks split scheduling into two layers that optimize in isolation: model routers pick a model from quality and cost signals while ignoring instance load, and serving load balancers optimize queues while ignoring quality. We present RouteBalance, a serving-aware scheduling layer that fuses both into a single online assignment over concrete model instances, jointly trading off quality, latency, and cost. A batched in-process predictor stack and dead-reckoned instance state keep the joint decision cheap on the request hot path ($\\approx$32 ms at 12 req/s). On a 13-insta","authors_text":"Evangelia Kalyvianaki, Wei Da","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.DC","submitted_at":"2026-06-16T14:02:34Z","title":"RouteBalance: Fused Model Routing and Load Balancing for Heterogeneous LLM Serving"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.17949","kind":"arxiv","version":1},"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:b687ee8424c58299650c3de38df15c3ffbe5e7f724bf2b457d2c88f7162d501b","target":"record","created_at":"2026-06-19T16:10:44Z","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":"8275f93eb0c8062d0b50e1c351de1bb1d63517a27fcac44d38a58f2b14da8c13","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.DC","submitted_at":"2026-06-16T14:02:34Z","title_canon_sha256":"bf1af85a692642e400b21f38fa200e92675da516f80ced4fb07756698b57cc78"},"schema_version":"1.0","source":{"id":"2606.17949","kind":"arxiv","version":1}},"canonical_sha256":"43808cb216ec306e8dcdf732ee91917fb565a8a7d1b92aaa189a5b8eb4125f27","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"43808cb216ec306e8dcdf732ee91917fb565a8a7d1b92aaa189a5b8eb4125f27","first_computed_at":"2026-06-19T16:10:44.310069Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-19T16:10:44.310069Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"Zl8xXlhB9ziAO61EkUyaBFQPQXbN3kANWeo7YFQWOjhRbr43/3dzAGWc9cVpoN/xp6UiSZV5y972z6LNmSSRCQ==","signature_status":"signed_v1","signed_at":"2026-06-19T16:10:44.310431Z","signed_message":"canonical_sha256_bytes"},"source_id":"2606.17949","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:b687ee8424c58299650c3de38df15c3ffbe5e7f724bf2b457d2c88f7162d501b","sha256:008d45d65792600dec285b3839676971275e2b4a8ae32d9b82eb80215c7489fc"],"state_sha256":"482ac1f8b358b5420569b3161cdd69e5f9fe66f2c2c37b708ae93b1821cff8f5"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"rXFGs0WWLlw/AcdxbqwoPKcqdPCUsL3Te4Z+dkUQAWt8OiL7K9L1u8hPyxOgeRm1l9CYk06fR4Fma17CKTH/AQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-04T20:50:39.120118Z","bundle_sha256":"aa9616b1bc22a35b5cc9894db1ace3b8929e942557a86b11b9be902555a827e5"}}