pith:3UNJHOKX
Taming Request Imbalance: SLO-Aware Scheduling for Disaggregated LLM Inference
Kairos improves TTFT SLO attainment by up to 24% and decode throughput by 19% in disaggregated LLM inference.
arxiv:2605.02329 v2 · 2026-05-04 · cs.DC
Add to your LaTeX paper
\usepackage{pith}
\pithnumber{3UNJHOKX3ZIP752Q66AOLWZAQP}
Prints a linked badge after your title and injects PDF metadata. Compiles on arXiv. Learn more · Embed verified badge
Record completeness
Claims
Experimental results demonstrate that, compared with state-of-the-art baselines, Kairos improves TTFT SLO attainment by up to 23.9%, TPOT SLO attainment by up to 27.1%, end-to-end SLO attainment by up to 33.8%, and decode throughput by up to 19.3%.
That the prediction of prefill completion times is accurate enough to make good scheduling decisions and that the chosen online serving dataset accurately reflects production request patterns.
Kairos improves SLO attainment and throughput in LLM serving by adapting to request length imbalance with priority scheduling and adaptive batching.
Formal links
Receipt and verification
| First computed | 2026-05-26T02:05:09.868224Z |
|---|---|
| Builder | pith-number-builder-2026-05-17-v1 |
| Signature | Pith Ed25519
(pith-v1-2026-05) · public key |
| Schema | pith-number/v1.0 |
Canonical hash
dd1a93b957de50fff750f780e5db2083dded13054e9a0bbd29450f62ee6d2bfe
Aliases
· · · · ·Agent API
Verify this Pith Number yourself
curl -sH 'Accept: application/ld+json' https://pith.science/pith/3UNJHOKX3ZIP752Q66AOLWZAQP \
| jq -c '.canonical_record' \
| python3 -c "import sys,json,hashlib; b=json.dumps(json.loads(sys.stdin.read()), sort_keys=True, separators=(',',':'), ensure_ascii=False).encode(); print(hashlib.sha256(b).hexdigest())"
# expect: dd1a93b957de50fff750f780e5db2083dded13054e9a0bbd29450f62ee6d2bfe
Canonical record JSON
{
"metadata": {
"abstract_canon_sha256": "2f81bf5886885711cede8487d5194cbe0e1b50b1eb314e43eb0b87ea60eee50e",
"cross_cats_sorted": [],
"license": "http://creativecommons.org/licenses/by/4.0/",
"primary_cat": "cs.DC",
"submitted_at": "2026-05-04T08:29:47Z",
"title_canon_sha256": "a1c0c0c6b87927f1d32f0c9a2fad3a0392d4fb5f372a6d8dc5d74e4adc743c03"
},
"schema_version": "1.0",
"source": {
"id": "2605.02329",
"kind": "arxiv",
"version": 2
}
}