mlsanityData to 7 Oct 2026

The Last Mile Is the File: OfficeEditBench for Preservation-Aware Office Editing

arXiv 2609.34469 ↗First posted 28 Sep 2026cs.SE cs.CLcited by 0

Not on the map. Why?

Abstract (arXiv)

A small Office edit creates two obligations: propagate every required update and leave protected state untouched. Updating too little leaves dependencies inconsistent; updating too much changes content the user did not authorize. We introduce OfficeEditBench, a 170-task benchmark for change-scoped maintenance of spreadsheets, presentations, and documents. Task contracts specify required updates, protected state, native structures, and applicable interaction requirements. Across 510 archived task-system outcomes from WorkBuddy, Doubao, and Codex, we distinguish file delivery, target completion, and verifier-defined acceptance. Hard package-valid delivery ranges from 92% to 100%, yet no selected output satisfies the complete contract. Case analysis highlights why local correctness is insufficient: an updated value can lose its generating formula, a revised rule can fail to reach related conclusions, and a new deadline can omit a retained prerequisite. These mechanisms connect artifact-level checks to the continued maintainability of Office files. We analyze maintenance failures while distinguishing frozen automatic verdicts from human acceptability. OfficeEditBench provides a testbed for completing required changes while preserving the logic and scope of existing work.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 16 of its 16 references that are in the map. First posted on arXiv 28 Sep 2026.

Builds on

16 of 16 references.

References · 16

  1. 2023SWE-bench: Can Language Models Resolve Real-World GitHub Issues?cited by 3,193
  2. 2023WebArena: A Realistic Web Environment for Building Autonomous Agentscited by 1,908
  3. 2018Datasheets for Datasetscited by 1,371
  4. 2024τ-bench: A Benchmark for Tool-Agent-User Interaction in Real-World Domainscited by 1,280
  5. 2024OSWorld: Benchmarking Multimodal Agents for Open-Ended Tasks in Real Computer Environmentscited by 1,134
  6. 2024MLE-bench: Evaluating Machine Learning Agents on Machine Learning Engineeringcited by 429
  7. 2024LiveBench: A Challenging, Contamination-Limited LLM Benchmarkcited by 242
  8. 2024ScienceAgentBench: Toward Rigorous Assessment of Language Agents for Data-Driven Scientific Discoverycited by 208
  9. 2024SpreadsheetBench: Towards Challenging Real World Spreadsheet Manipulationcited by 100
  10. 2024WorkArena++: Towards Compositional Planning and Reasoning-based Common Knowledge Work Taskscited by 81
  11. 2024OfficeBench: Benchmarking Language Agents across Multiple Applications for Office Automationcited by 66
  12. 2025PPTArena: A Benchmark for PowerPoint Editingcited by 9
4 more
  1. 2026SpreadsheetBench 2: Evaluating Agents on End-to-End Business Spreadsheet Workflowscited by 7
  2. 2026PPT-Eval: A Benchmark for Computer-Use Agents on PowerPoint Taskscited by 6
  3. 2026OmegaUse-OfficeVal: Benchmarking LLM Agents on Long-Horizon Office-Suite Tasks with Economic Groundingcited by 1
  4. 2026EditPPT: Faithful Long-Deck Slide Editing via Structured Tool-Using Multi-Agent with Dual-Modal Validatorscited by 1

Built on by

Not cited yet.