paper-with-me

홈 › Papers

Investigating the Benefits of Free-Form Rationales

2022-05-25 · Jiao Sun, Swabha Swayamdipta, Jonathan May, Xuezhe Ma

Free-form rationales aim to aid model interpretability by supplying the background knowledge that can help understand model decisions. Crowdsourced rationales are provided for commonsense QA instances in popular datasets such as CoS-E and ECQA, but their utility remains under-investigated. We present human studies which show that ECQA rationales indeed provide additional background information to understand a decision, while over 88% of CoS-E rationales do not. Inspired by this finding, we ask: can the additional context provided by free-form rationales benefit models, similar to human users? We investigate the utility of rationales as an additional source of supervision, by varying the quantity and quality of rationales during training. After controlling for instances where rationales leak the correct answer while not providing additional background knowledge, we find that incorporating only 5% of rationales during training can boost model performance by 47.22% for CoS-E and 57.14% for ECQA during inference. Moreover, we also show that rationale quality matters: compared to crowdsourced rationales, T5-generated rationales provide not only weaker supervision to models, but are also not helpful for humans in aiding model interpretability.

📄 PDF Abstract BibTeX arXiv:2206.11083

Code (0)

등록된 구현이 없습니다.

Tasks

Form

Similar Papers 제목 키워드 기반

Investigating the Benefits of Free-Form Rationales

2022-01-16 · ACL ARR January 2022 1 · Anonymous

Free-form rationales aim to aid model interpretability by supplying the background knowledge that can help understand model decisions. Crowdsourced rationales are provided for commonsense QA instances in popular datasets…

Form

Natural Language Rationales with Full-Stack Visual Reasoning: From Pixels to Semantic Frames to Commonsense Graphs

2020-10-15 · Findings of the Association for Computational Linguistics 2020 · Ana Marasović, Chandra Bhagavatula, Jae Sung Park, Ronan Le Bras 외

Natural language rationales could provide intuitive, higher-level explanations that are easily understandable by humans, complementing the more broadly studied lower-level explanations based on gradients or attention wei…

Language ModelingLanguage ModellingNatural Language InferenceObject Recognition+5

FRAME: Evaluating Rationale-Label Consistency Metrics for Free-Text Rationales

2022-07-02 · Aaron Chan, Shaoliang Nie, Liang Tan, Xiaochang Peng 외

Following how humans communicate, free-text rationales aim to use natural language to explain neural language model (LM) behavior. However, free-text rationales' unconstrained nature makes them prone to hallucination, so…

HallucinationLanguage Modellingtext-classificationText Classification

Measuring Association Between Labels and Free-Text Rationales

2020-10-24 · EMNLP 2021 11 · Sarah Wiegreffe, Ana Marasović, Noah A. Smith

In interpretable NLP, we require faithful rationales that reflect the model's decision-making process for an explained instance. While prior work focuses on extractive rationales (a subset of the input words), we investi…

Decision MakingFeature ImportanceNatural Language InferenceQuestion Answering

TrClaim-19: The First Collection for Turkish Check-Worthy Claim Detection with Annotator Rationales

2020-11-01 · CONLL 2020 · Yavuz Selim Kartal, Mucahid Kutlu

Massive misinformation spread over Internet has many negative impacts on our lives. While spreading a claim is easy, investigating its veracity is hard and time consuming, Therefore, we urgently need systems to help huma…

Misinformation