A Benchmark and Framework for Evaluating Next Action Predictions in Spreadsheets

Hugging Face Daily Papers Papers
benchmark spreadsheet next-action-prediction llm fine-tuning evaluation

Summary

This paper introduces a benchmark for predicting spreadsheet user actions, addressing challenges in edit history availability and complex action spaces through manual curation and online evaluation methodology.

Predictive code completion greatly accelerates how quickly developers work. In spreadsheets, despite being much more common, such auto-completion features are virtually non-existent. To address this gap, we introduce a benchmark for systems that observe a sequence of user actions in a spreadsheet and predict future actions. Two challenges are (1) the absence of edit histories in public spreadsheet corpora and (2) the complex space of spreadsheet actions (spatial, temporal, composite). To address (1), we manually curate 52 sequences of 12K actions that recreate spreadsheets from public corpora, seeded by parametrized heuristics and LLM refinement. To address (2), we propose an online evaluation that expects a prediction after each user action, accepts or rejects that prediction, updates the future actions upon acceptance, and repeats this until the target spreadsheet is obtained. We use multiple baseline predictors (including zero-shot LLMs, fine-tuned SLMs, and classical models) and analyze different properties that our benchmark teaches us, including but not limited to: properties of saved actions and false positives, efficiency, effect of user profiles, effect of triggers, and effect of context.
Original Article
View Cached Full Text

Cached at: 06/18/26, 03:58 PM

Paper page - A Benchmark and Framework for Evaluating Next Action Predictions in Spreadsheets

Source: https://huggingface.co/papers/2606.13802

Abstract

A benchmark for predicting spreadsheet user actions is introduced, addressing challenges in edit history availability and complex action spaces through manual curation and online evaluation methodology.

Predictivecodecompletiongreatlyaccelerateshowquicklydeveloperswork.Inspreadsheets,despitebeingmuchmorecommon,suchauto-completionfeaturesarevirtuallynon-existent.Toaddressthisgap,weintroduceabenchmarkforsystemsthatobserveasequenceofuseractionsinaspreadsheetandpredictfutureactions.Twochallengesare(1)theabsenceofedithistoriesinpublicspreadsheetcorporaand(2)thecomplexspaceofspreadsheetactions(spatial,temporal,composite).Toaddress(1),wemanuallycurate52sequencesof12Kactionsthatrecreatespreadsheetsfrompubliccorpora,seededbyparametrizedheuristicsandLLMrefinement.Toaddress(2),weproposeanonlineevaluationthatexpectsapredictionaftereachuseraction,acceptsorrejectsthatprediction,updatesthefutureactionsuponacceptance,andrepeatsthisuntilthetargetspreadsheetisobtained.Weusemultiplebaselinepredictors(includingzero-shotLLMs,fine-tunedSLMs,andclassicalmodels)andanalyzedifferentpropertiesthatourbenchmarkteachesus,includingbutnotlimitedto:propertiesofsavedactionsandfalsepositives,efficiency,effectofuserprofiles,effectoftriggers,andeffectofcontext.

View arXiv pageView PDFProject pageGitHub0Add to collection

Get this paper in your agent:

hf papers read 2606\.13802

Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash

Models citing this paper0

No model linking this paper

Cite arxiv.org/abs/2606.13802 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2606.13802 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2606.13802 in a Space README.md to link it from this page.

Collections including this paper0

No Collection including this paper

Add this paper to acollectionto link it from this page.

Similar Articles

Back to the Future: A workbook time machine for spread sheet creation benchmarks

arXiv cs.AI

This paper introduces the workbook time machine, a pipeline that automatically creates benchmarks for evaluating language models on creating derived spreadsheet objects like formulas, charts, pivot tables, and conditional formatting. The authors produce WTM-Corpus and a curated 150-task benchmark WTM-Bench, then evaluate spreadsheet agents across artifact types, step complexity, and instruction granularity.

PreAct-Bench: Benchmarking Predictive Monitoring in LLMs

arXiv cs.LG

PreAct-Bench is a benchmark of 1,000 paired ethical and unethical action trajectories across five domains, designed to evaluate the ability of LLMs to predict harmful outcomes from partial trajectories (predictive monitoring). Results show that while humans perform well, current LLMs struggle, highlighting the need for future-oriented risk reasoning.

BehaviorBench: Benchmarking Foundation Models for Behavioral Science Tasks

arXiv cs.CL

This paper introduces BehaviorBench, a comprehensive benchmark for evaluating foundation models on behavioral science tasks including behavior prediction, strategic decision-making, subject-trait inference, and behavioral knowledge application. It also presents Be.FM-1.5, a fine-tuned model that achieves strong distributional alignment, highlighting the gap between general-purpose and behaviorally adapted models.

From Heuristics to Analytics: Forecasting Effort and Progress in Online Learning

arXiv cs.LG

This paper introduces engagement forecasting for intelligent tutoring systems, predicting weekly minutes practiced and new skills mastered using interaction logs from 425 middle-school students. Feature-based models reduce error by 22-33% over heuristic baselines, offering explainable patterns for tutor-learner goal setting.