Skip to content
OpenTrain AIFor AI Companies

HFEPX · Eval paper review

Diagnosing Translated Benchmarks: An Automated Quality Assurance Study of the EU20 Benchmark Suite

Klaudia Thellmann, Bernhard Stadler, Michael Färber

Published

Apr 2, 2026

Citations

0

Trust level

Low

Usefulness score

15/100 (Low)

Extraction confidence

55% (Moderate)

Derived from extracted protocol signals and abstract evidence.

Rater population

Not reported

Signals refreshed

Apr 2, 2026

Open arXiv Find implementation Shortlist (0) RSS feed

Tags

Automatic Metrics

Should you rely on this paper?

This paper is adjacent to HFEPX scope and is best used for background context, not as a primary protocol reference.

Use this as background context only. Do not make protocol decisions from this page alone.

Best use

Background context only

Use if you need

A benchmark-and-metrics comparison anchor.

What to verify

Validate the exact study setup in the full paper before operational use.

Main weakness

This paper looks adjacent to evaluation work, but not like a strong protocol reference.

Human feedback signal
Not explicit
Not explicit in abstract metadata
Evaluation signal
Detected
Eval setup described
Usefulness for eval research
15/100
Adjacent candidate

Treat as adjacent context, not a core eval-method reference.

If you are doing eval pipeline work, start here

Human Eval Papers LLM-as-Judge Papers Pairwise Preference Papers Tool-Use Evaluation Search

Abstract

Machine-translated benchmark datasets reduce costs and offer scale, but noise, loss of structure, and uneven quality weaken confidence. What matters is not merely whether we can translate, but also whether we can measure and verify translation reliability at scale. We study translation quality in the EU20 benchmark suite, which comprises five established benchmarks translated into 20 languages, via a three-step automated quality assurance approach: (i) a structural corpus audit with targeted fixes; (ii) quality profiling using a neural metric (COMET, reference-free and reference-based) with translation service comparisons (DeepL / ChatGPT / Google); and (iii) an LLM-based span-level translation error landscape. Trends are consistent: datasets with lower COMET scores exhibit a higher share of accuracy/mistranslation errors at span level (notably HellaSwag; ARC is comparatively clean). Reference-based COMET on MMLU against human-edited samples points in the same direction. We release cleaned/corrected versions of the EU20 datasets, and code for reproducibility. In sum, automated quality assurance offers practical, scalable indicators that help prioritize review -- complementing, not replacing, human gold standards.

What we could verify

These are the protocol signals we could actually recover from the available paper metadata. Use them to decide whether this paper is worth deeper reading.

Human Feedback Types

missing

None explicit

No explicit feedback protocol extracted.

"Machine-translated benchmark datasets reduce costs and offer scale, but noise, loss of structure, and uneven quality weaken confidence."

Evaluation Modes

strong

Automatic Metrics

Includes extracted eval setup.

"Machine-translated benchmark datasets reduce costs and offer scale, but noise, loss of structure, and uneven quality weaken confidence."

Quality Controls

strong

Calibration, Gold Questions

Calibration/adjudication style controls detected.

"Machine-translated benchmark datasets reduce costs and offer scale, but noise, loss of structure, and uneven quality weaken confidence."

Benchmarks / Datasets

strong

MMLU, HellaSwag

Useful for quick benchmark comparison.

"Trends are consistent: datasets with lower COMET scores exhibit a higher share of accuracy/mistranslation errors at span level (notably HellaSwag; ARC is comparatively clean)."

Reported Metrics

strong

Accuracy

Useful for evaluation criteria comparison.

"Trends are consistent: datasets with lower COMET scores exhibit a higher share of accuracy/mistranslation errors at span level (notably HellaSwag; ARC is comparatively clean)."

Benchmarks and datasets

MMLUHellaSwag

Reported metrics

accuracy
Human feedback details
Uses human feedback
No
Feedback types
None
Rater population
Not reported
Expertise required
Coding, Multilingual
Evaluation details
Evaluation modes
Automatic Metrics
Agentic eval
None
Quality controls
Calibration, Gold Questions
Evidence quality
Low
Use this page as
Background context only

Research brief

Metadata summary

Machine-translated benchmark datasets reduce costs and offer scale, but noise, loss of structure, and uneven quality weaken confidence.

Based on abstract + metadata only. Check the source paper before making high-confidence protocol decisions.

Key takeaways

  • Machine-translated benchmark datasets reduce costs and offer scale, but noise, loss of structure, and uneven quality weaken confidence.
  • What matters is not merely whether we can translate, but also whether we can measure and verify translation reliability at scale.
  • We study translation quality in the EU20 benchmark suite, which comprises five established benchmarks translated into 20 languages, via a three-step automated quality assurance approach: (i) a structural corpus audit with targeted fixes; (ii) quality profiling using a neural metric (COMET, reference-free and reference-based) with translation service comparisons (DeepL / ChatGPT / Google); and (iii) an LLM-based span-level translation error landscape.

Researcher actions

  • Compare this paper against others mentioning MMLU.
  • Validate inferred eval signals (Automatic metrics) against the full paper.
  • Use related-paper links to find stronger protocol-specific references.

Caveats

  • Generated from abstract + metadata only; no PDF parsing.
  • Signals below are heuristic and may miss details reported outside the abstract.

Recommended queries

MMLU Accuracy

Contribution summary

  • Machine-translated benchmark datasets reduce costs and offer scale, but noise, loss of structure, and uneven quality weaken confidence.
  • We study translation quality in the EU20 benchmark suite, which comprises five established benchmarks translated into 20 languages, via a three-step automated quality assurance approach: (i) a structural corpus audit with targeted fixes;…
  • Reference-based COMET on MMLU against human-edited samples points in the same direction.

Why it matters for eval

  • Machine-translated benchmark datasets reduce costs and offer scale, but noise, loss of structure, and uneven quality weaken confidence.
  • We study translation quality in the EU20 benchmark suite, which comprises five established benchmarks translated into 20 languages, via a three-step automated quality assurance approach: (i) a structural corpus audit with targeted fixes;…

Researcher checklist

  • Human feedback protocol is explicit

    No explicit human feedback protocol detected.

  • Evaluation mode is explicit

    Detected: Automatic Metrics

  • Quality control reporting appears

    Detected: Calibration, Gold Questions

  • Benchmark or dataset anchors are present

    Detected: MMLU, HellaSwag

  • Metric reporting is present

    Detected: accuracy