TaskEval: Synthesised Evaluation for Foundation-Model Tasks

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Widanapathiranage, Dilani, Barnett, Scott, Kurniawan, Stefanus, Takerngsaksiri, Wannita
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909945479299072
author Widanapathiranage, Dilani
Barnett, Scott
Kurniawan, Stefanus
Takerngsaksiri, Wannita
author_facet Widanapathiranage, Dilani
Barnett, Scott
Kurniawan, Stefanus
Takerngsaksiri, Wannita
contents Hallucinations are a key concern when creating applications that rely on Foundation models (FMs). Understanding where and how these subtle failures occur in an application relies on evaluation methods known as \textit{evals}. Prior work focuses on defining new eval methods or benchmark datasets for specific tasks. However, neither helps a software team with a task-specific FM application when there is no metric or dataset. The demand for both automated approaches and deep integration of human insight makes this a challenging problem. We address this gap by proposing an approach to synthesise a FM task-specific evaluator program that provides automation and a custom UI for capturing feedback. The core novelty of our approach lies in: (1) a task-agnostic meta-model that captures properties of any FM task, (2) an interaction protocol for efficient use of human feedback, and (3) an eval synthesiser that selects or generates an appropriate set of evals. We implement our approach in \toolname and demonstrate the concept on two diverse FM tasks: chart data extraction and document question answering. A preliminary evaluation on the quality of our selected evals shows 93\% and 90\% accuracy respectively. Our research tackles a growing problem facing engineering teams, how to evaluate and review outputs from FM tasks.
format Preprint
id arxiv_https___arxiv_org_abs_2512_04442
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle TaskEval: Synthesised Evaluation for Foundation-Model Tasks
Widanapathiranage, Dilani
Barnett, Scott
Kurniawan, Stefanus
Takerngsaksiri, Wannita
Artificial Intelligence
Software Engineering
Hallucinations are a key concern when creating applications that rely on Foundation models (FMs). Understanding where and how these subtle failures occur in an application relies on evaluation methods known as \textit{evals}. Prior work focuses on defining new eval methods or benchmark datasets for specific tasks. However, neither helps a software team with a task-specific FM application when there is no metric or dataset. The demand for both automated approaches and deep integration of human insight makes this a challenging problem. We address this gap by proposing an approach to synthesise a FM task-specific evaluator program that provides automation and a custom UI for capturing feedback. The core novelty of our approach lies in: (1) a task-agnostic meta-model that captures properties of any FM task, (2) an interaction protocol for efficient use of human feedback, and (3) an eval synthesiser that selects or generates an appropriate set of evals. We implement our approach in \toolname and demonstrate the concept on two diverse FM tasks: chart data extraction and document question answering. A preliminary evaluation on the quality of our selected evals shows 93\% and 90\% accuracy respectively. Our research tackles a growing problem facing engineering teams, how to evaluate and review outputs from FM tasks.
title TaskEval: Synthesised Evaluation for Foundation-Model Tasks
topic Artificial Intelligence
Software Engineering
url https://arxiv.org/abs/2512.04442