# Structural analysis of an all-purpose question answering model

Source: [Structural analysis of an all-purpose question answering model](<https://www.sota2.com/research/paper/68fa69bd1e209131cea23ec6>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2025-10-27T06:06:39.643Z

Authors: Vincent Micheli, Quentin Heinrich, Fran\\c{c}ois Fleuret, Wacim Belblidia

Published: 2021-04-13T09:20:44.000Z

## Abstract

Attention is a key component of the now ubiquitous pre-trained language models. By learning to focus on relevant pieces of information, these Transformer-based architectures have proven capable of tackling several tasks at once and sometimes even surpass their single-task counterparts. To better understand this phenomenon, we conduct a structural analysis of a new all-purpose question answering model that we introduce. Surprisingly, this model retains single-task performance even in the absence of a strong transfer effect between tasks. Through attention head importance scoring, we observe that attention heads specialize in a particular task and that some heads are more conducive to learning than others in both the multi-task and single-task settings.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2104.06045>)

[PDF](<https://arxiv.org/pdf/2104.06045.pdf>)