Skip to content
OpenTrain AIFor AI Companies
← Back to explorer

Multilingual Prompt Localization for Agent-as-a-Judge: Language and Backbone Sensitivity in Requirement-Level Evaluation

Alhasan Mahmood, Samir Abdaljalil, Hasan Kurban · Apr 6, 2026 · Citations: 0

Llm As Judge
Open arXiv Find Implementation RSS feed Shortlist (0)

How to use this page

Low trust

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

Best use

Background context only

What to verify

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

Evidence quality

Low

Derived from extracted protocol signals and abstract evidence.

Abstract

Evaluation language is typically treated as a fixed English default in agentic code benchmarks, yet we show that changing the judge's language can invert backbone rankings. We localize the Agent-as-a-Judge prompt stack to five typologically diverse languages (English, Arabic, Turkish, Chinese, Hindi) and evaluate 55 DevAI development tasks across three developer-agent frameworks and six judge backbones, totaling 4950 judge runs. The central finding is that backbone and language interact: GPT-4o achieves the highest satisfaction in English (44.72\%), while Gemini leads in Arabic (51.72\%, $p<0.001$ vs.\ GPT-4o) and Hindi (53.22\%). No single backbone dominates across all languages, and inter-backbone agreement on individual requirement judgments is modest (Fleiss' $κ\leq 0.231$). A controlled ablation further shows that localizing judge-side instructions, not just benchmark content, can be decisive: Hindi satisfaction drops from 42.8\% to 23.2\% under partial localization. These results indicate that language should be treated as an explicit evaluation variable in agentic benchmarks. Full requirement-level judgments and runtime statistics are released for reproducibility.

Abstract-only analysis — low confidence

All signals on this page are inferred from the abstract only and may be inaccurate. Do not use this page as a primary protocol reference.

  • This paper looks adjacent to evaluation work, but not like a strong protocol reference.
  • The available metadata is too thin to trust this as a primary source.
Human Eval Hub LLM-as-Judge Hub Pairwise Preference Hub

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.

Best use

Background context only

Use if you need

A secondary eval reference to pair with stronger protocol papers.

Main weakness

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

Trust level

Low

Usefulness score

17/100 • Low

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

Human Feedback Signal

Not explicit in abstract metadata

Evaluation Signal

Detected

Usefulness for eval research

Adjacent candidate

Extraction confidence 45%

If you are doing eval pipeline work, start here:

Human Eval Hub LLM-as-Judge Hub Pairwise Preference Hub Tool-Use Eval Hub

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.

"Evaluation language is typically treated as a fixed English default in agentic code benchmarks, yet we show that changing the judge's language can invert backbone rankings."

Evaluation Modes

partial

Llm As Judge

Includes extracted eval setup.

"Evaluation language is typically treated as a fixed English default in agentic code benchmarks, yet we show that changing the judge's language can invert backbone rankings."

Quality Controls

partial

Inter Annotator Agreement Reported

Calibration/adjudication style controls detected.

"Evaluation language is typically treated as a fixed English default in agentic code benchmarks, yet we show that changing the judge's language can invert backbone rankings."

Benchmarks / Datasets

missing

Not extracted

No benchmark anchors detected.

"Evaluation language is typically treated as a fixed English default in agentic code benchmarks, yet we show that changing the judge's language can invert backbone rankings."

Reported Metrics

partial

Agreement

Useful for evaluation criteria comparison.

"No single backbone dominates across all languages, and inter-backbone agreement on individual requirement judgments is modest (Fleiss' $κ\leq 0.231$)."

Human Feedback Details

  • Uses human feedback: No
  • Feedback types: None
  • Rater population: Not reported
  • Unit of annotation: Ranking (inferred)
  • Expertise required: Coding, Multilingual

Evaluation Details

  • Evaluation modes: Llm As Judge
  • Agentic eval: None
  • Quality controls: Inter Annotator Agreement Reported
  • Evidence quality: Low
  • Use this page as: Background context only

Protocol And Measurement Signals

Benchmarks / Datasets

No benchmark or dataset names were extracted from the available abstract.

Reported Metrics

agreement

Research Brief

Metadata summary

Evaluation language is typically treated as a fixed English default in agentic code benchmarks, yet we show that changing the judge's language can invert backbone rankings.

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

Key Takeaways

  • Evaluation language is typically treated as a fixed English default in agentic code benchmarks, yet we show that changing the judge's language can invert backbone rankings.
  • We localize the Agent-as-a-Judge prompt stack to five typologically diverse languages (English, Arabic, Turkish, Chinese, Hindi) and evaluate 55 DevAI development tasks across three developer-agent frameworks and six judge backbones, totaling 4950 judge runs.
  • The central finding is that backbone and language interact: GPT-4o achieves the highest satisfaction in English (44.72\%), while Gemini leads in Arabic (51.72\%, $p<0.001$ vs.\ GPT-4o) and Hindi (53.22\%).

Researcher Actions

  • Compare this paper against nearby papers in the same arXiv category before using it for protocol decisions.
  • Check the full text for explicit evaluation design choices (raters, protocol, and metrics).
  • 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

Agreement / Kappa

Research Summary

Contribution Summary

  • Evaluation language is typically treated as a fixed English default in agentic code benchmarks, yet we show that changing the judge's language can invert backbone rankings.
  • We localize the Agent-as-a-Judge prompt stack to five typologically diverse languages (English, Arabic, Turkish, Chinese, Hindi) and evaluate 55 DevAI development tasks across three developer-agent frameworks and six judge backbones,…
  • A controlled ablation further shows that localizing judge-side instructions, not just benchmark content, can be decisive: Hindi satisfaction drops from 42.8\% to 23.2\% under partial localization.

Why It Matters For Eval

  • Evaluation language is typically treated as a fixed English default in agentic code benchmarks, yet we show that changing the judge's language can invert backbone rankings.
  • We localize the Agent-as-a-Judge prompt stack to five typologically diverse languages (English, Arabic, Turkish, Chinese, Hindi) and evaluate 55 DevAI development tasks across three developer-agent frameworks and six judge backbones,…

Researcher Checklist

  • Gap: Human feedback protocol is explicit

    No explicit human feedback protocol detected.

  • Pass: Evaluation mode is explicit

    Detected: Llm As Judge

  • Pass: Quality control reporting appears

    Detected: Inter Annotator Agreement Reported

  • Gap: Benchmark or dataset anchors are present

    No benchmark/dataset anchor extracted from abstract.

  • Pass: Metric reporting is present

    Detected: agreement

Related Papers

Papers are ranked by protocol overlap, extraction signal alignment, and semantic proximity.