# Models in the Loop: Aiding Crowdworkers with Generative Annotation Assistants

arXiv: https://arxiv.org/abs/2112.09062 · HTML: https://mlsanity.com/paper/2112.09062/

## Key facts

- Cited by 31 papers in the map (data through 14 Oct 2026).
- Builds on 34 of its 42 references that are in the map.
- First posted on arXiv 16 Dec 2021.

## Builds on (most cited first, 30 of 34)

- [BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehension](https://mlsanity.com/paper/1910.13461/) (2019)
- [SQuAD: 100,000+ Questions for Machine Comprehension of Text](https://mlsanity.com/paper/1606.05250/) (2016)
- [fairseq: A Fast, Extensible Toolkit for Sequence Modeling](https://mlsanity.com/paper/1904.01038/) (2019)
- [DROP: A Reading Comprehension Benchmark Requiring Discrete Reasoning Over Paragraphs](https://mlsanity.com/paper/1903.00161/) (2019)
- [Right for the Wrong Reasons: Diagnosing Syntactic Heuristics in Natural Language Inference](https://mlsanity.com/paper/1902.01007/) (2019)
- [Adversarial NLI: A New Benchmark for Natural Language Understanding](https://mlsanity.com/paper/1910.14599/) (2019)
- [Annotation Artifacts in Natural Language Inference Data](https://mlsanity.com/paper/1803.02324/) (2018)
- [KILT: a Benchmark for Knowledge Intensive Language Tasks](https://mlsanity.com/paper/2009.02252/) (2020)
- [Adversarial Examples for Evaluating Reading Comprehension Systems](https://mlsanity.com/paper/1707.07328/) (2017)
- [Dynabench: Rethinking Benchmarking in NLP](https://mlsanity.com/paper/2104.14337/) (2021)
- [Polyjuice: Generating Counterfactuals for Explaining, Evaluating, and Improving Models](https://mlsanity.com/paper/2101.00288/) (2021)
- [What Will it Take to Fix Benchmarking in Natural Language Understanding?](https://mlsanity.com/paper/2104.02145/) (2021)
- [PAQ: 65 Million Probably-Asked Questions and What You Can Do With Them](https://mlsanity.com/paper/2102.07033/) (2021)
- [Learning to Ask: Neural Question Generation for Reading Comprehension](https://mlsanity.com/paper/1705.00106/) (2017)
- [MRQA 2019 Shared Task: Evaluating Generalization in Reading Comprehension](https://mlsanity.com/paper/1910.09753/) (2019)
- [Build it Break it Fix it for Dialogue Safety: Robustness from Adversarial Human Attack](https://mlsanity.com/paper/1908.06083/) (2019)
- [Are We Modeling the Task or the Annotator? An Investigation of Annotator Bias in Natural Language Understanding Datasets](https://mlsanity.com/paper/1908.07898/) (2019)
- [Synthetic QA Corpora Generation with Roundtrip Consistency](https://mlsanity.com/paper/1906.05416/) (2019)
- [Beat the AI: Investigating Adversarial Human Annotation for Reading Comprehension](https://mlsanity.com/paper/2002.00293/) (2020)
- [Training Question Answering Models From Synthetic Data](https://mlsanity.com/paper/2002.09599/) (2020)
- [Learning and Evaluating General Linguistic Intelligence](https://mlsanity.com/paper/1901.11373/) (2019)
- [Improving Question Answering Model Robustness with Synthetic Adversarial Data Generation](https://mlsanity.com/paper/2104.08678/) (2021)
- [A Thorough Examination of the CNN/Daily Mail Reading Comprehension Task](https://mlsanity.com/paper/1606.02858/) (2016)
- [Tailor: Generating and Perturbing Text with Semantic Controls](https://mlsanity.com/paper/2107.07150/) (2021)
- [DynaSent: A Dynamic Benchmark for Sentiment Analysis](https://mlsanity.com/paper/2012.15349/) (2020)
- [Comparing Test Sets with Item Response Theory](https://mlsanity.com/paper/2106.00840/) (2021)
- [Harvesting Paragraph-Level Question-Answer Pairs from Wikipedia](https://mlsanity.com/paper/1805.05942/) (2018)
- [The Effect of Different Writing Tasks on Linguistic Style: A Case Study of the ROC Story Cloze Task](https://mlsanity.com/paper/1702.01841/) (2017)
- [Making Neural QA as Simple as Possible but not Simpler](https://mlsanity.com/paper/1703.04816/) (2017)
- [On the Efficacy of Adversarial Data Collection for Question Answering: Results from a Large-Scale Randomized Study](https://mlsanity.com/paper/2106.00872/) (2021)

## Built on by (most cited first, 30 of 31)

- [Jailbreaking Black Box Large Language Models in Twenty Queries](https://mlsanity.com/paper/2310.08419/) (2023)
- [Red Teaming Language Models with Language Models](https://mlsanity.com/paper/2202.03286/) (2022)
- [Discovering Language Model Behaviors with Model-Written Evaluations](https://mlsanity.com/paper/2212.09251/) (2022)
- [WANLI: Worker and AI Collaboration for Natural Language Inference Dataset Creation](https://mlsanity.com/paper/2201.05955/) (2022)
- [CoAnnotating: Uncertainty-Guided Work Allocation between Human and Large Language Models for Data Annotation](https://mlsanity.com/paper/2310.15638/) (2023)
- [Adversarial Training for High-Stakes Reliability](https://mlsanity.com/paper/2205.01663/) (2022)
- [PURR: Efficiently Editing Language Model Hallucinations by Denoising Language Model Corruptions](https://mlsanity.com/paper/2305.14908/) (2023)
- [Cofca: A Step-Wise Counterfactual Multi-hop QA benchmark](https://mlsanity.com/paper/2402.11924/) (2024)
- [CORE: A Retrieve-then-Edit Framework for Counterfactual Data Generation](https://mlsanity.com/paper/2210.04873/) (2022)
- [Sociocultural Norm Similarities and Differences via Situational Alignment and Explainable Textual Entailment](https://mlsanity.com/paper/2305.14492/) (2023)
- [ScatterShot: Interactive In-context Example Curation for Text Transformation](https://mlsanity.com/paper/2302.07346/) (2023)
- [Adversarial Nibbler: An Open Red-Teaming Method for Identifying Diverse Harms in Text-to-Image Generation](https://mlsanity.com/paper/2403.12075/) (2024)
- [Evaluating Visual and Cultural Interpretation: The K-Viscuit Benchmark with Human-VLM Collaboration](https://mlsanity.com/paper/2406.16469/) (2024)
- [Dynatask: A Framework for Creating Dynamic AI Benchmark Tasks](https://mlsanity.com/paper/2204.01906/) (2022)
- [A Theory of Dynamic Benchmarks](https://mlsanity.com/paper/2210.03165/) (2022)
- [Improving Adversarial Data Collection by Supporting Annotators: Lessons from GAHD, a German Hate Speech Dataset](https://mlsanity.com/paper/2403.19559/) (2024)
- [mCSQA: Multilingual Commonsense Reasoning Dataset with Unified Creation Strategy by Language Models and Humans](https://mlsanity.com/paper/2406.04215/) (2024)
- [On the Limitations of Simulating Active Learning](https://mlsanity.com/paper/2305.13342/) (2023)
- [CoCoLoFa: A Dataset of News Comments with Common Logical Fallacies Written by LLM-Assisted Crowds](https://mlsanity.com/paper/2410.03457/) (2024)
- [Adversarial Nibbler: A Data-Centric Challenge for Improving the Safety of Text-to-Image Models](https://mlsanity.com/paper/2305.14384/) (2023)
- [ICLEF: In-Context Learning with Expert Feedback for Explainable Style Transfer](https://mlsanity.com/paper/2309.08583/) (2023)
- [Cascading Biases: Investigating the Effect of Heuristic Annotation Strategies on Data and Models](https://mlsanity.com/paper/2210.13439/) (2022)
- [Variability Matters : Evaluating inter-rater variability in histopathology for robust cell detection](https://mlsanity.com/paper/2210.05175/) (2022)
- [CoAct: Co-Active LLM Preference Learning with Human-AI Synergy](https://mlsanity.com/paper/2604.17501/) (2026)
- [A Human-ML Collaboration Framework for Improving Video Content Reviews](https://mlsanity.com/paper/2210.09500/) (2022)
- [Out-of-Distribution Generalization in Text Classification: Past, Present, and Future](https://mlsanity.com/paper/2305.14104/) (2023)
- [On Degrees of Freedom in Defining and Testing Natural Language Understanding](https://mlsanity.com/paper/2305.15130/) (2023)
- [Power-up! What Can Generative Models Do for Human Computation Workflows?](https://mlsanity.com/paper/2307.02243/) (2023)
- [OLAF: Towards Robust LLM-Based Annotation Framework in Empirical Software Engineering](https://mlsanity.com/paper/2512.15979/) (2025)
- [DataCI: A Platform for Data-Centric AI on Streaming Data](https://mlsanity.com/paper/2306.15538/) (2023)

## Abstract (arXiv)

In Dynamic Adversarial Data Collection (DADC), human annotators are tasked with finding examples that models struggle to predict correctly. Models trained on DADC-collected training data have been shown to be more robust in adversarial and out-of-domain settings, and are considerably harder for humans to fool. However, DADC is more time-consuming than traditional data collection and thus more costly per annotated example. In this work, we examine whether we can maintain the advantages of DADC, without incurring the additional cost. To that end, we introduce Generative Annotation Assistants (GAAs), generator-in-the-loop models that provide real-time suggestions that annotators can either approve, modify, or reject entirely. We collect training datasets in twenty experimental settings and perform a detailed analysis of this approach for the task of extractive question answering (QA) for both standard and adversarial data collection. We demonstrate that GAAs provide significant efficiency benefits with over a 30% annotation speed-up, while leading to over a 5x improvement in model fooling rates. In addition, we find that using GAA-assisted training data leads to higher downstream model performance on a variety of question answering tasks over adversarial data collection.