{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:6RVCV2MFWNQ5VJU7QJYEQND3RJ","short_pith_number":"pith:6RVCV2MF","schema_version":"1.0","canonical_sha256":"f46a2ae985b361daa69f827048347b8a619434601dea0425f50967c6fa11c699","source":{"kind":"arxiv","id":"2505.19634","version":4},"attestation_state":"computed","paper":{"title":"Faster and Better LLMs via Latency-Aware Test-Time Scaling","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Haoli Bai, Lei Zhu, Lu Hou, Shiming Xiang, Tianyu Zhang, Wulong Liu, Xianzhi Yu, Zili Wang","submitted_at":"2025-05-26T07:51:30Z","abstract_excerpt":"Test-Time Scaling (TTS) has proven effective in improving the performance of Large Language Models (LLMs) during inference. However, existing research has overlooked the efficiency of TTS from a latency-sensitive perspective. Through a latency-aware evaluation of representative TTS methods, we demonstrate that a compute-optimal TTS does not always result in the lowest latency in scenarios where latency is critical. To address this gap and achieve latency-optimal TTS, we propose two key approaches by optimizing the concurrency configurations: (1) branch-wise parallelism, which leverages multipl"},"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":"2505.19634","kind":"arxiv","version":4},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-26T07:51:30Z","cross_cats_sorted":[],"title_canon_sha256":"081c2c4b1aa089a7c2bb51f66393f5678ade2e434879cb86120fbcce8de29f71","abstract_canon_sha256":"fb9b87b628a25ea97ec432f897eedb67fb1e2141b2da112629728fb6bf844feb"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T12:09:44.508284Z","signature_b64":"grr4SEtcgyEDwp7OMs3ChsXFn2sTzSmf3LwRKvsFvARJEDXc/hXCiCL+r2zENB/3VS1ZCeeElT91GSdAqCn8Bg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f46a2ae985b361daa69f827048347b8a619434601dea0425f50967c6fa11c699","last_reissued_at":"2026-07-05T12:09:44.507811Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T12:09:44.507811Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Faster and Better LLMs via Latency-Aware Test-Time Scaling","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Haoli Bai, Lei Zhu, Lu Hou, Shiming Xiang, Tianyu Zhang, Wulong Liu, Xianzhi Yu, Zili Wang","submitted_at":"2025-05-26T07:51:30Z","abstract_excerpt":"Test-Time Scaling (TTS) has proven effective in improving the performance of Large Language Models (LLMs) during inference. However, existing research has overlooked the efficiency of TTS from a latency-sensitive perspective. Through a latency-aware evaluation of representative TTS methods, we demonstrate that a compute-optimal TTS does not always result in the lowest latency in scenarios where latency is critical. To address this gap and achieve latency-optimal TTS, we propose two key approaches by optimizing the concurrency configurations: (1) branch-wise parallelism, which leverages multipl"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.19634","kind":"arxiv","version":4},"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/2505.19634/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":"2505.19634","created_at":"2026-07-05T12:09:44.507885+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.19634v4","created_at":"2026-07-05T12:09:44.507885+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.19634","created_at":"2026-07-05T12:09:44.507885+00:00"},{"alias_kind":"pith_short_12","alias_value":"6RVCV2MFWNQ5","created_at":"2026-07-05T12:09:44.507885+00:00"},{"alias_kind":"pith_short_16","alias_value":"6RVCV2MFWNQ5VJU7","created_at":"2026-07-05T12:09:44.507885+00:00"},{"alias_kind":"pith_short_8","alias_value":"6RVCV2MF","created_at":"2026-07-05T12:09:44.507885+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2608.08447","citing_title":"Hidden Language Consistency Phenomena in Reasoning LLMs","ref_index":285,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/6RVCV2MFWNQ5VJU7QJYEQND3RJ","json":"https://pith.science/pith/6RVCV2MFWNQ5VJU7QJYEQND3RJ.json","graph_json":"https://pith.science/api/pith-number/6RVCV2MFWNQ5VJU7QJYEQND3RJ/graph.json","events_json":"https://pith.science/api/pith-number/6RVCV2MFWNQ5VJU7QJYEQND3RJ/events.json","paper":"https://pith.science/paper/6RVCV2MF"},"agent_actions":{"view_html":"https://pith.science/pith/6RVCV2MFWNQ5VJU7QJYEQND3RJ","download_json":"https://pith.science/pith/6RVCV2MFWNQ5VJU7QJYEQND3RJ.json","view_paper":"https://pith.science/paper/6RVCV2MF","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.19634&json=true","fetch_graph":"https://pith.science/api/pith-number/6RVCV2MFWNQ5VJU7QJYEQND3RJ/graph.json","fetch_events":"https://pith.science/api/pith-number/6RVCV2MFWNQ5VJU7QJYEQND3RJ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/6RVCV2MFWNQ5VJU7QJYEQND3RJ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/6RVCV2MFWNQ5VJU7QJYEQND3RJ/action/storage_attestation","attest_author":"https://pith.science/pith/6RVCV2MFWNQ5VJU7QJYEQND3RJ/action/author_attestation","sign_citation":"https://pith.science/pith/6RVCV2MFWNQ5VJU7QJYEQND3RJ/action/citation_signature","submit_replication":"https://pith.science/pith/6RVCV2MFWNQ5VJU7QJYEQND3RJ/action/replication_record"}},"created_at":"2026-07-05T12:09:44.507885+00:00","updated_at":"2026-07-05T12:09:44.507885+00:00"}