1 code implementation • 16 Nov 2023 • Yixiao Song, Kalpesh Krishna, Rajesh Bhatt, Kevin Gimpel, Mohit Iyyer
To address this gap, we propose the task of grammar error explanation, where a system needs to provide one-sentence explanations for each grammatical error in a pair of erroneous and corrected sentences.
1 code implementation • 29 May 2023 • Fangyuan Xu, Yixiao Song, Mohit Iyyer, Eunsol Choi
We present a careful analysis of experts' evaluation, which focuses on new aspects such as the comprehensiveness of the answer.
no code implementations • 24 May 2023 • Shufan Wang, Yixiao Song, Andrew Drozdov, Aparna Garimella, Varun Manjunatha, Mohit Iyyer
Digging deeper, we find that interpolating with a retrieval distribution actually increases perplexity compared to a baseline Transformer LM for the majority of tokens in the WikiText-103 test set, even though the overall perplexity is lower due to a smaller number of tokens for which perplexity dramatically decreases after interpolation.
1 code implementation • NeurIPS 2023 • Kalpesh Krishna, Yixiao Song, Marzena Karpinska, John Wieting, Mohit Iyyer
To increase the robustness of AI-generated text detection to paraphrase attacks, we introduce a simple defense that relies on retrieving semantically-similar generations and must be maintained by a language model API provider.
1 code implementation • 25 Oct 2022 • Marzena Karpinska, Nishant Raj, Katherine Thai, Yixiao Song, Ankita Gupta, Mohit Iyyer
While machine translation evaluation metrics based on string overlap (e. g., BLEU) have their limitations, their computations are transparent: the BLEU score assigned to a particular candidate translation can be traced back to the presence or absence of certain words.
1 code implementation • 21 Oct 2022 • Yixiao Song, Kalpesh Krishna, Rajesh Bhatt, Mohit Iyyer
To understand what kinds of linguistic knowledge are encoded by pretrained Chinese language models (LMs), we introduce the benchmark of Sino LINGuistics (SLING), which consists of 38K minimal sentence pairs in Mandarin Chinese grouped into 9 high-level linguistic phenomena.