{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:XKNCSZMSDP6ZQZXTYOG7JJZQQU","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":"faa9a0297bc7f252a301293c2fe069d6198c36ac360931d795254fb3745d77ac","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2024-05-28T01:53:26Z","title_canon_sha256":"e1c2ac49ad97f314d10790d2db4a9f7a0c7a19c431c679a014d98f1335b7f0d3"},"schema_version":"1.0","source":{"id":"2405.17741","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2405.17741","created_at":"2026-07-05T08:24:10Z"},{"alias_kind":"arxiv_version","alias_value":"2405.17741v1","created_at":"2026-07-05T08:24:10Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2405.17741","created_at":"2026-07-05T08:24:10Z"},{"alias_kind":"pith_short_12","alias_value":"XKNCSZMSDP6Z","created_at":"2026-07-05T08:24:10Z"},{"alias_kind":"pith_short_16","alias_value":"XKNCSZMSDP6ZQZXT","created_at":"2026-07-05T08:24:10Z"},{"alias_kind":"pith_short_8","alias_value":"XKNCSZMS","created_at":"2026-07-05T08:24:10Z"}],"graph_snapshots":[{"event_id":"sha256:101154b5032e19dca6c7e942b9327d2d6bf8fadc76a7f70a73fefe2067c1bfa3","target":"graph","created_at":"2026-07-05T08:24:10Z","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/2405.17741/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Recent literature has found that an effective method to customize or further improve large language models (LLMs) is to add dynamic adapters, such as low-rank adapters (LoRA) with Mixture-of-Experts (MoE) structures. Though such dynamic adapters incur modest computational complexity, they surprisingly lead to huge inference latency overhead, slowing down the decoding speed by 2.5+ times. In this paper, we analyze the fine-grained costs of the dynamic adapters and find that the fragmented CUDA kernel calls are the root cause. Therefore, we propose LoRA-Switch, a system-algorithm co-designed arc","authors_text":"Linghe Kong, Qingfeng He, Qingtian Feng, Qiyang Li, Rui Kong, Weijun Wang, Xinyu Fang, Yazhu Dong, Yuanchun Li, Yunxin Liu","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2024-05-28T01:53:26Z","title":"LoRA-Switch: Boosting the Efficiency of Dynamic LLM Adapters via System-Algorithm Co-design"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2405.17741","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:c1d022946fe127d6548206381474af1bd47aa45c04101d376ae0ba0e7fc6a4a2","target":"record","created_at":"2026-07-05T08:24:10Z","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":"faa9a0297bc7f252a301293c2fe069d6198c36ac360931d795254fb3745d77ac","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2024-05-28T01:53:26Z","title_canon_sha256":"e1c2ac49ad97f314d10790d2db4a9f7a0c7a19c431c679a014d98f1335b7f0d3"},"schema_version":"1.0","source":{"id":"2405.17741","kind":"arxiv","version":1}},"canonical_sha256":"ba9a2965921bfd9866f3c38df4a7308517c929054503639651c62ccc85dc3f08","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"ba9a2965921bfd9866f3c38df4a7308517c929054503639651c62ccc85dc3f08","first_computed_at":"2026-07-05T08:24:10.545694Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:24:10.545694Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"nSqozg8fgw4kOA2QJg4QzHuxu3Ff/9ZphLTIY9uUmUV9PfUQJwpsQBeaIedPjV/JFZGGpdwn/gZRFLRcVaR3Aw==","signature_status":"signed_v1","signed_at":"2026-07-05T08:24:10.546138Z","signed_message":"canonical_sha256_bytes"},"source_id":"2405.17741","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:c1d022946fe127d6548206381474af1bd47aa45c04101d376ae0ba0e7fc6a4a2","sha256:101154b5032e19dca6c7e942b9327d2d6bf8fadc76a7f70a73fefe2067c1bfa3"],"state_sha256":"1adef47d1be6766c3cb5ad9d9bb4e6d258720556b3a2a4eef819ddd7b49b01a7"}