Log In Sign Up

Holistic Evaluation of Language Models

by   Percy Liang, et al.

Language models (LMs) are becoming the foundation for almost all major language technologies, but their capabilities, limitations, and risks are not well understood. We present Holistic Evaluation of Language Models (HELM) to improve the transparency of language models. First, we taxonomize the vast space of potential scenarios (i.e. use cases) and metrics (i.e. desiderata) that are of interest for LMs. Then we select a broad subset based on coverage and feasibility, noting what's missing or underrepresented (e.g. question answering for neglected English dialects, metrics for trustworthiness). Second, we adopt a multi-metric approach: We measure 7 metrics (accuracy, calibration, robustness, fairness, bias, toxicity, and efficiency) for each of 16 core scenarios when possible (87.5 accuracy don't fall to the wayside, and that trade-offs are clearly exposed. We also perform 7 targeted evaluations, based on 26 targeted scenarios, to analyze specific aspects (e.g. reasoning, disinformation). Third, we conduct a large-scale evaluation of 30 prominent language models (spanning open, limited-access, and closed models) on all 42 scenarios, 21 of which were not previously used in mainstream LM evaluation. Prior to HELM, models on average were evaluated on just 17.9 models not sharing a single scenario in common. We improve this to 96.0 all 30 models have been densely benchmarked on the same core scenarios and metrics under standardized conditions. Our evaluation surfaces 25 top-level findings. For full transparency, we release all raw model prompts and completions publicly for further analysis, as well as a general modular toolkit. We intend for HELM to be a living benchmark for the community, continuously updated with new scenarios, metrics, and models.


On the Intrinsic and Extrinsic Fairness Evaluation Metrics for Contextualized Language Representations

Multiple metrics have been introduced to measure fairness in various nat...

Internet-augmented language models through few-shot prompting for open-domain question answering

In this work, we aim to capitalize on the unique few-shot capabilities o...

Refining Targeted Syntactic Evaluation of Language Models

Targeted syntactic evaluation of subject-verb number agreement in Englis...

Generative Language Models for Paragraph-Level Question Generation

Powerful generative models have led to recent progress in question gener...

Measuring Fairness with Biased Rulers: A Survey on Quantifying Biases in Pretrained Language Models

An increasing awareness of biased patterns in natural language processin...

Large Language Models Encode Clinical Knowledge

Large language models (LLMs) have demonstrated impressive capabilities i...

Prompting GPT-3 To Be Reliable

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

Code Repositories


Holistic Evaluation of Language Models (HELM), a framework to increase the transparency of language models (

view repo