# Do not let the history haunt you -- Mitigating Compounding Errors in Conversational Question Answering

arXiv: https://arxiv.org/abs/2005.05754 · HTML: https://mlsanity.com/paper/2005.05754/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 12 May 2020.

## Built on by (most cited first, 6 of 6)

- [Learning to Relate to Previous Turns in Conversational Search](https://mlsanity.com/paper/2306.02553/) (2023)
- [Ditch the Gold Standard: Re-evaluating Conversational Question Answering](https://mlsanity.com/paper/2112.08812/) (2021)
- [Interactive Question Answering Systems: Literature Review](https://mlsanity.com/paper/2209.01621/) (2022)
- [Beyond Static Evaluation: A Dynamic Approach to Assessing AI Assistants' API Invocation Capabilities](https://mlsanity.com/paper/2403.11128/) (2024)
- [A Minimal Approach for Natural Language Action Space in Text-based Games](https://mlsanity.com/paper/2305.04082/) (2023)
- [Realistic Conversational Question Answering with Answer Selection based on Calibrated Confidence and Uncertainty Measurement](https://mlsanity.com/paper/2302.05137/) (2023)

## Abstract (arXiv)

The Conversational Question Answering (CoQA) task involves answering a sequence of inter-related conversational questions about a contextual paragraph. Although existing approaches employ human-written ground-truth answers for answering conversational questions at test time, in a realistic scenario, the CoQA model will not have any access to ground-truth answers for the previous questions, compelling the model to rely upon its own previously predicted answers for answering the subsequent questions. In this paper, we find that compounding errors occur when using previously predicted answers at test time, significantly lowering the performance of CoQA systems. To solve this problem, we propose a sampling strategy that dynamically selects between target answers and model predictions during training, thereby closely simulating the situation at test time. Further, we analyse the severity of this phenomena as a function of the question type, conversation length and domain type.