Pith.

Operating Systems · arXiv:2611.08638 v1 · submitted 5 Nov 2026

MORTIS: Compiler-Guided KV-Cache Management for LLM Agent Programs

Xiaoyu Ma, Jayanaka L. Dantanarayana, Savini Kashmira, Baichuan Li, Lingjia Tang, Jason Mars

The paper says

“Our evaluation shows that MORTIS achieves up to 1.99x job completion time (JCT) speedup and 1.97x throughput improvement over SGLang, with geometric-mean speedups of 1.45x and 1.49x, respectively.”

from the abstract · picked out as the main claim by a classifier

Pith says

Pith has not reviewed this paper.

Get a Pith Review, $5

Reviews

Reviews by people

No person has reviewed this paper on Pith yet. A scientist listed on Pith can review it for a fee, and their review is published here whatever it concludes.

Request a human review

How Pith checks papers · The full audit of this paper

The paper

Abstract

Existing agent serving systems divide each prompt into a reusable prefix and an opaque suffix, managing the KV-cache at this coarse granularity without distinguishing the components within each part. We observe that prompt components differ in which LLM calls share their values and how long those values remain reusable, a property we term intra-prompt heterogeneity. Our key insight is that, in frameworks that generate prompts from typed LLM function calls, these components correspond to program variables, allowing compiler analysis to recover their sharing patterns and lifetimes. We present MORTIS, a compiler-serving co-design that combines this analysis with runtime observations to improve KV-cache management. MORTIS reorders prompt components to expose additional prefix reuse and uses their predicted reuse and lifetimes to guide cache retention, eviction, and prefetching. Our evaluation shows that MORTIS achieves up to 1.99x job completion time (JCT) speedup and 1.97x throughput improvement over SGLang, with geometric-mean speedups of 1.45x and 1.49x, respectively.

The paper's own sentences, arranged. Each label is drawn from a closed vocabulary; the number beside it is the probability the classifier gave that label. No sentence in this record was written by a model.

abstract compiled · 2026-11-09 · no language-model review

How the main claim is stated: asserted 1.00

What the abstract itself states, as probabilities: claims first 0.07 · states limitation 0.08 · human participants 0.02 · controlled comparison 0.93 · randomized assignment 0.01 · states quantitative result 0.99 · compares with prior results 0.97 · mentions code or data availability 0.03

The abstract, sentence by sentence

This table labels the abstract only (arXiv listing abstract v1). The number is the probability the classifier gave the label.

SentenceRoleThe authors' own workStates the main conclusionStrength
Existing agent serving systems divide each prompt into a reusable prefix and an opaque suffix, managing the KV-cache at this coarse granularity without distinguishing the components within each part. background 0.93 0.05 0.02 asserted 1.00
We observe that prompt components differ in which LLM calls share their values and how long those values remain reusable, a property we term intra-prompt heterogeneity. gap 0.43 0.90 0.04 asserted 1.00
Our key insight is that, in frameworks that generate prompts from typed LLM function calls, these components correspond to program variables, allowing compiler analysis to recover their sharing patterns and lifetimes. contribution 0.49 0.70 0.14 asserted 0.98
We present MORTIS, a compiler-serving co-design that combines this analysis with runtime observations to improve KV-cache management. contribution 0.98 0.89 0.13 asserted 0.99
MORTIS reorders prompt components to expose additional prefix reuse and uses their predicted reuse and lifetimes to guide cache retention, eviction, and prefetching. method 0.97 0.89 0.15 asserted 1.00
Our evaluation shows that MORTIS achieves up to 1.99x job completion time (JCT) speedup and 1.97x throughput improvement over SGLang, with geometric-mean speedups of 1.45x and 1.49x, respectively. result 1.00 0.97 0.64 asserted 1.00

What kind of paper the classifier reports this is: method or tool 1.00 · evidence from observation or experiment 0.99

What was examined, and what was not

Sources read

  • arXiv listing abstract v1

Not examined

  • full text
  • figures and tables
  • references and citing sentences
  • relations to other papers

Levels run

  • P abstract

Not run

  • S sections
  • T body sentences
  • C citations
  • R cross-paper relations

How the labels were made

Compiled by classifier-1.13.0 via TypeSafe API, question bank abstract-p1-20260923, 2026-11-09: 1 request, 5,057 input tokens.

Every quoted sentence on this record is the authors'. Every label is from Pith's closed vocabulary, with the probability shown beside it. This record as JSON

Cited by 1 Pith paper

Reviewed papers on Pith that cite this one. The full citation record

  1. Agent-Native by Design: A Programming Language for Agentic Applications cs.PL · Nov 2026

Challenges and discussion

Know something about this paper? Leave a comment.

No comments yet.

Sign in to comment