Get our free extension to see links to code for papers anywhere online!

Add to Chrome

Add to Firefox

Picture for Sewon Min

Sewon Min

Revisiting Calibration for Question Answering

May 25, 2022
Chenglei Si, Chen Zhao, Sewon Min, Jordan Boyd-Graber

Figure 1 for Revisiting Calibration for Question Answering

Figure 2 for Revisiting Calibration for Question Answering

Figure 3 for Revisiting Calibration for Question Answering

Figure 4 for Revisiting Calibration for Question Answering

* Preprint; Feedback is welcome

Access Paper or Ask Questions

Rethinking the Role of Demonstrations: What Makes In-Context Learning Work?

Feb 25, 2022
Sewon Min, Xinxi Lyu, Ari Holtzman, Mikel Artetxe, Mike Lewis, Hannaneh Hajishirzi, Luke Zettlemoyer

Figure 1 for Rethinking the Role of Demonstrations: What Makes In-Context Learning Work?

Figure 2 for Rethinking the Role of Demonstrations: What Makes In-Context Learning Work?

Figure 3 for Rethinking the Role of Demonstrations: What Makes In-Context Learning Work?

Figure 4 for Rethinking the Role of Demonstrations: What Makes In-Context Learning Work?

* 14 pages; 11 figures; Code available at https://github.com/Alrope123/rethinking-demonstrations

Access Paper or Ask Questions

MetaICL: Learning to Learn In Context

Oct 29, 2021
Sewon Min, Mike Lewis, Luke Zettlemoyer, Hannaneh Hajishirzi

Figure 1 for MetaICL: Learning to Learn In Context

Figure 2 for MetaICL: Learning to Learn In Context

Figure 3 for MetaICL: Learning to Learn In Context

Figure 4 for MetaICL: Learning to Learn In Context

* 18 pages (9 pages for the main paper, 9 pages for references and appendices). 1 figure. Code available at https://github.com/facebookresearch/MetaICL

Access Paper or Ask Questions

Noisy Channel Language Model Prompting for Few-Shot Text Classification

Aug 15, 2021
Sewon Min, Mike Lewis, Hannaneh Hajishirzi, Luke Zettlemoyer

Figure 1 for Noisy Channel Language Model Prompting for Few-Shot Text Classification

Figure 2 for Noisy Channel Language Model Prompting for Few-Shot Text Classification

Figure 3 for Noisy Channel Language Model Prompting for Few-Shot Text Classification

Figure 4 for Noisy Channel Language Model Prompting for Few-Shot Text Classification

* 15 pages, 6 figures

Access Paper or Ask Questions

FaVIQ: FAct Verification from Information-seeking Questions

Jul 05, 2021
Jungsoo Park, Sewon Min, Jaewoo Kang, Luke Zettlemoyer, Hannaneh Hajishirzi

Figure 1 for FaVIQ: FAct Verification from Information-seeking Questions

Figure 2 for FaVIQ: FAct Verification from Information-seeking Questions

Figure 3 for FaVIQ: FAct Verification from Information-seeking Questions

Figure 4 for FaVIQ: FAct Verification from Information-seeking Questions

* 12 pages, 3 figures; Data & Code available at https://faviq.github.io

Access Paper or Ask Questions

NeurIPS 2020 EfficientQA Competition: Systems, Analyses and Lessons Learned

Jan 01, 2021
Sewon Min, Jordan Boyd-Graber, Chris Alberti, Danqi Chen, Eunsol Choi, Michael Collins, Kelvin Guu, Hannaneh Hajishirzi, Kenton Lee, Jennimaria Palomaki, Colin Raffel, Adam Roberts, Tom Kwiatkowski, Patrick Lewis, Yuxiang Wu, Heinrich Küttler, Linqing Liu, Pasquale Minervini, Pontus Stenetorp, Sebastian Riedel, Sohee Yang, Minjoon Seo, Gautier Izacard, Fabio Petroni, Lucas Hosseini, Nicola De Cao, Edouard Grave, Ikuya Yamada, Sonse Shimaoka, Masatoshi Suzuki, Shumpei Miyawaki, Shun Sato, Ryo Takahashi, Jun Suzuki, Martin Fajcik, Martin Docekal, Karel Ondrej, Pavel Smrz, Hao Cheng, Yelong Shen, Xiaodong Liu, Pengcheng He, Weizhu Chen, Jianfeng Gao, Barlas Oguz, Xilun Chen, Vladimir Karpukhin, Stan Peshterliev, Dmytro Okhonko, Michael Schlichtkrull, Sonal Gupta, Yashar Mehdad, Wen-tau Yih

Figure 1 for NeurIPS 2020 EfficientQA Competition: Systems, Analyses and Lessons Learned

Figure 2 for NeurIPS 2020 EfficientQA Competition: Systems, Analyses and Lessons Learned

Figure 3 for NeurIPS 2020 EfficientQA Competition: Systems, Analyses and Lessons Learned

Figure 4 for NeurIPS 2020 EfficientQA Competition: Systems, Analyses and Lessons Learned

* 26 pages

Access Paper or Ask Questions