Pith. sign in

REVIEW 2 cited by

Dialectal Toxicity Detection: Evaluating LLM-as-a-Judge Consistency Across Language Varieties

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2411.10954 v1 pith:TUJBK2LI submitted 2024-11-17 cs.CL

Dialectal Toxicity Detection: Evaluating LLM-as-a-Judge Consistency Across Language Varieties

classification cs.CL
keywords dialectalllmstoxicityconsistencyacrossareadetectionlanguage
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
Share X Bluesky LinkedIn Reddit HN
read the original abstract

There has been little systematic study on how dialectal differences affect toxicity detection by modern LLMs. Furthermore, although using LLMs as evaluators ("LLM-as-a-judge") is a growing research area, their sensitivity to dialectal nuances is still underexplored and requires more focused attention. In this paper, we address these gaps through a comprehensive toxicity evaluation of LLMs across diverse dialects. We create a multi-dialect dataset through synthetic transformations and human-assisted translations, covering 10 language clusters and 60 varieties. We then evaluated three LLMs on their ability to assess toxicity across multilingual, dialectal, and LLM-human consistency. Our findings show that LLMs are sensitive in handling both multilingual and dialectal variations. However, if we have to rank the consistency, the weakest area is LLM-human agreement, followed by dialectal consistency. Code repository: \url{https://github.com/ffaisal93/dialect_toxicity_llm_judge}

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 2 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Measurement Without Validity: The Compounding Reliability Problem in Agentic AI Evaluation

    cs.AI 2026-08 conditional novelty 5.0

    Agentic evaluation pipelines can retain as little as a third of the intended measurement signal because task, simulation, and judgment errors multiply, while most published inter-rater reliability reporting is structu...

  2. Measurement Without Validity: The Compounding Reliability Problem in Agentic AI Evaluation

    cs.AI 2026-08 reject novelty 5.0

    Agentic AI evaluation validity is bounded by the product of task-generation, simulator, and judge reliability, leaving most current automated benchmarks with less than 30% valid signal.