Measuring and Mitigating Toxicity in Large Language Models: A Comprehensive Replication Study
This replication study evaluates DExperts, an inference-time toxicity mitigation technique that requires no retraining. It achieved 100% safety on explicit toxicity benchmarks but dropped to 98.5% on implicit hate speech, with a ~10x latency penalty.
[2605.14087] Measuring and Mitigating Toxicity in Large Language Models: A Comprehensive Replication Study
[Submitted on 13 May 2026]
Title:Measuring and Mitigating Toxicity in Large Language Models: A Comprehensive Replication Study
View a PDF of the paper titled Measuring and Mitigating Toxicity in Large Language Models: A Comprehensive Replication Study, by Mokshit Surana and 1 other authors
View PDF HTML (experimental)
Abstract:Large Language Models (LLMs), when trained on web-scale corpora, inherently absorb toxic patterns from their training data. This leads to ``toxic degeneration'' where even innocuous prompts can trigger harmful outputs. This phenomenon poses significant risks for real-world deployments. Thus, necessitating effective mitigation strategies that should maintain model utility while ensuring safety. In this comprehensive replication study, we evaluate the efficacy of \textbf{DExperts} (Decoding-time Experts), which is an inference-time mitigation technique that steers generation without requiring model retraining. We structured our research into three systematic phases: (1) establishing baseline toxicity measurements using \textbf{RealToxicityPrompts} on standard GPT-2 models; then (2) implementing and evaluating DExperts to mitigate explicit toxicity; and finally (3) stress-testing the method against implicit hate speech using the adversarial \textbf{ToxiGen} dataset. Our empirical results confirm that while DExperts achieves near-perfect safety rates (100\%) on explicit toxicity benchmarks, it exhibits brittleness against adversarial, implicit hate speech, with safety rates dropping to 98.5\%. Furthermore, we quantify a critical trade-off. The method introduces a $\sim$10x latency penalty (from 0.2s to 2.0s per generation), posing challenges for real-time deployment scenarios. This study contributes to the growing body of work on AI safety by highlighting the robustness gap between explicit and implicit toxicity mitigation. We emphasize the need for more sophisticated approaches that generalize across diverse hate speech patterns without prohibitive computational costs.
Subjects:
Computation and Language (cs.CL); Machine Learning (cs.LG)
Cite as: arXiv:2605.14087 [cs.CL]
(or arXiv:2605.14087v1 [cs.CL] for this version)
https://doi.org/10.48550/arXiv.2605.14087
arXiv-issued DOI via DataCite (pending registration)
Submission history
From: Mokshit Surana [view email] [v1] Wed, 13 May 2026 20:12:30 UTC (1,797 KB)
Full-text links:
Access Paper:
View a PDF of the paper titled Measuring and Mitigating Toxicity in Large Language Models: A Comprehensive Replication Study, by Mokshit Surana and 1 other authors
View PDF
HTML (experimental)
TeX Source
view license
Current browse context:
cs.CL
new | recent | 2026-05
Change to browse by:
cs cs.LG
References & Citations
NASA ADS
Google Scholar
Semantic Scholar
Loading...
Data provided by:
Bibliographic Tools
Bibliographic and Citation Tools
Bibliographic Explorer Toggle
Bibliographic Explorer (What is the Explorer?)
Connected Papers Toggle
Connected Papers (What is Connected Papers?)
Litmaps Toggle
Litmaps (What is Litmaps?)
scite.ai Toggle
scite Smart Citations (What are Smart Citations?)
Code, Data, Media
Code, Data and Media Associated with this Article
alphaXiv Toggle
alphaXiv (What is alphaXiv?)
Links to Code Toggle
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub Toggle
DagsHub (What is DagsHub?)
GotitPub Toggle
Gotit.pub (What is GotitPub?)
Huggingface Toggle
Hugging Face (What is Huggingface?)
ScienceCast Toggle
ScienceCast (What is ScienceCast?)
Demos
Demos
Replicate Toggle
Replicate (What is Replicate?)
Spaces Toggle
Hugging Face Spaces (What is Spaces?)
Spaces Toggle
TXYZ.AI (What is TXYZ.AI?)
Related Papers
Recommenders and Search Tools
Link to Influence Flower
Influence Flower (What are Influence Flowers?)
Core recommender toggle
CORE Recommender (What is CORE?)
Author
Venue
Institution
Topic
About arXivLabs
arXivLabs: experimental projects with community collaborators
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?)