Picture for Mrinmaya Sachan

Mrinmaya Sachan

Automatic Generation of Socratic Subquestions for Teaching Math Word Problems

Nov 23, 2022
Kumar Shridhar, Jakub Macina, Mennatallah El-Assady, Tanmay Sinha, Manu Kapur, Mrinmaya Sachan

Figure 1 for Automatic Generation of Socratic Subquestions for Teaching Math Word Problems

Figure 2 for Automatic Generation of Socratic Subquestions for Teaching Math Word Problems

Figure 3 for Automatic Generation of Socratic Subquestions for Teaching Math Word Problems

Figure 4 for Automatic Generation of Socratic Subquestions for Teaching Math Word Problems

* Kumar Shridhar and Jakub Macina contributed equally to this work. Accepted at the 2022 Conference on Empirical Methods in Natural Language Processing (EMNLP 2022). Code available: https://github.com/eth-nlped/scaffolding-generation

Access Paper or Ask Questions

Autoregressive Structured Prediction with Language Models

Nov 17, 2022
Tianyu Liu, Yuchen Jiang, Nicholas Monath, Ryan Cotterell, Mrinmaya Sachan

Figure 1 for Autoregressive Structured Prediction with Language Models

Figure 2 for Autoregressive Structured Prediction with Language Models

Figure 3 for Autoregressive Structured Prediction with Language Models

Figure 4 for Autoregressive Structured Prediction with Language Models

* EMNLP 2022 (findings)

Access Paper or Ask Questions

Beyond prompting: Making Pre-trained Language Models Better Zero-shot Learners by Clustering Representations

Oct 29, 2022
Yu Fei, Ping Nie, Zhao Meng, Roger Wattenhofer, Mrinmaya Sachan

Figure 1 for Beyond prompting: Making Pre-trained Language Models Better Zero-shot Learners by Clustering Representations

Figure 2 for Beyond prompting: Making Pre-trained Language Models Better Zero-shot Learners by Clustering Representations

Figure 3 for Beyond prompting: Making Pre-trained Language Models Better Zero-shot Learners by Clustering Representations

Figure 4 for Beyond prompting: Making Pre-trained Language Models Better Zero-shot Learners by Clustering Representations

* Accepted to EMNLP 2022

Access Paper or Ask Questions

Differentially Private Language Models for Secure Data Sharing

Oct 26, 2022
Justus Mattern, Zhijing Jin, Benjamin Weggenmann, Bernhard Schoelkopf, Mrinmaya Sachan

Figure 1 for Differentially Private Language Models for Secure Data Sharing

Figure 2 for Differentially Private Language Models for Secure Data Sharing

Figure 3 for Differentially Private Language Models for Secure Data Sharing

Figure 4 for Differentially Private Language Models for Secure Data Sharing

* Accepted at EMNLP 2022

Access Paper or Ask Questions

Adapters for Enhanced Modeling of Multilingual Knowledge and Text

Oct 26, 2022
Yifan Hou, Wenxiang Jiao, Meizhen Liu, Carl Allen, Zhaopeng Tu, Mrinmaya Sachan

Figure 1 for Adapters for Enhanced Modeling of Multilingual Knowledge and Text

Figure 2 for Adapters for Enhanced Modeling of Multilingual Knowledge and Text

Figure 3 for Adapters for Enhanced Modeling of Multilingual Knowledge and Text

Figure 4 for Adapters for Enhanced Modeling of Multilingual Knowledge and Text

* Our code, models, and data (e.g., integration corpus and extended datasets) are available: https://github.com/yifan-h/Multilingual_Space

Access Paper or Ask Questions