DeepAI AI Chat
Log In Sign Up

Characteristics of Harmful Text: Towards Rigorous Benchmarking of Language Models

by   Maribeth Rauh, et al.

Large language models produce human-like text that drive a growing number of applications. However, recent literature and, increasingly, real world observations, have demonstrated that these models can generate language that is toxic, biased, untruthful or otherwise harmful. Though work to evaluate language model harms is under way, translating foresight about which harms may arise into rigorous benchmarks is not straightforward. To facilitate this translation, we outline six ways of characterizing harmful text which merit explicit consideration when designing new benchmarks. We then use these characteristics as a lens to identify trends and gaps in existing benchmarks. Finally, we apply them in a case study of the Perspective API, a toxicity classifier that is widely used in harm benchmarks. Our characteristics provide one piece of the bridge that translates between foresight and effective evaluation.


page 6

page 26


Limits of Detecting Text Generated by Large-Scale Language Models

Some consider large-scale language models that can generate long and coh...

Prompting GPT-3 To Be Reliable

Large language models (LLMs) show impressive abilities via few-shot prom...

Language Model Evaluation Beyond Perplexity

We propose an alternate approach to quantifying how well language models...

Do Language Models Know the Way to Rome?

The global geometry of language models is important for a range of appli...

TimeLMs: Diachronic Language Models from Twitter

Despite its importance, the time variable has been largely neglected in ...

Efficient Training of Language Models to Fill in the Middle

We show that autoregressive language models can learn to infill text aft...

Curb Your Carbon Emissions: Benchmarking Carbon Emissions in Machine Translation

In recent times, there has been definitive progress in the field of NLP,...