1 code implementation • 7 Oct 2024 • Daoan Zhang, Guangchen Lan, Dong-Jun Han, Wenlin Yao, Xiaoman Pan, Hongming Zhang, Mingxiao Li, Pengcheng Chen, Yu Dong, Christopher Brinton, Jiebo Luo
To address the limitations of both on- and off-policy RLHF, we propose a preference optimization method that aligns DMs with preferences without relying on reward models or paired human-annotated data.
1 code implementation • 3 Oct 2024 • Zhaowei Wang, Hongming Zhang, Tianqing Fang, Ye Tian, Yue Yang, Kaixin Ma, Xiaoman Pan, Yangqiu Song, Dong Yu
In this paper, we study a new task of navigating to diverse target objects in a large number of scene types.
1 code implementation • 16 Sep 2024 • Hongming Zhang, Xiaoman Pan, Hongwei Wang, Kaixin Ma, Wenhao Yu, Dong Yu
Cognitive Kernel adopts a model-centric design.
1 code implementation • 18 Jun 2024 • Ruixin Hong, Hongming Zhang, Xiaoman Pan, Dong Yu, ChangShui Zhang
Abstract reasoning, the ability to reason from the abstract essence of a problem, serves as a key to generalization in human reasoning.
1 code implementation • 27 Feb 2024 • Xinran Zhao, Hongming Zhang, Xiaoman Pan, Wenlin Yao, Dong Yu, Tongshuang Wu, Jianshu Chen
For a LLM to be trustworthy, its confidence level should be well-calibrated with its actual performance.
2 code implementations • 15 Feb 2024 • Rui Yang, Xiaoman Pan, Feng Luo, Shuang Qiu, Han Zhong, Dong Yu, Jianshu Chen
We consider the problem of multi-objective alignment of foundation models with human preferences, which is a critical step towards helpful and harmless AI systems.
no code implementations • 14 Dec 2023 • Kaiqiang Song, Xiaoyang Wang, Sangwoo Cho, Xiaoman Pan, Dong Yu
This paper introduces a novel approach to enhance the capabilities of Large Language Models (LLMs) in processing and understanding extensive text sequences, a critical aspect in applications requiring deep comprehension and synthesis of large volumes of information.
no code implementations • 15 Nov 2023 • Wenhao Yu, Hongming Zhang, Xiaoman Pan, Kaixin Ma, Hongwei Wang, Dong Yu
In response to these challenges, we introduces Chain-of-Noting (CoN), a novel approach aimed at improving the robustness of RALMs in facing noisy, irrelevant documents and in handling unknown scenarios.
1 code implementation • 30 Sep 2023 • Xuansheng Wu, Wenlin Yao, Jianshu Chen, Xiaoman Pan, Xiaoyang Wang, Ninghao Liu, Dong Yu
In this work, we investigate how the instruction tuning adjusts pre-trained models with a focus on intrinsic changes.
1 code implementation • 15 Sep 2023 • Kaixin Ma, Hongming Zhang, Hongwei Wang, Xiaoman Pan, Wenhao Yu, Dong Yu
We evaluate our proposed LLM Agent with State-Space ExploRation (LASER) on both the WebShop task and amazon. com.
no code implementations • 1 Aug 2023 • Jiaao Chen, Xiaoman Pan, Dian Yu, Kaiqiang Song, Xiaoyang Wang, Dong Yu, Jianshu Chen
We investigate how to elicit compositional generalization capabilities in large language models (LLMs).
Ranked #34 on
Math Word Problem Solving
on MATH
no code implementations • 16 Jul 2023 • Zhenwen Liang, Dian Yu, Xiaoman Pan, Wenlin Yao, Qingkai Zeng, Xiangliang Zhang, Dong Yu
Our approach uniquely considers the various annotation formats as different "views" and leverages them in training the model.
1 code implementation • 24 May 2023 • Keming Lu, Xiaoman Pan, Kaiqiang Song, Hongming Zhang, Dong Yu, Jianshu Chen
In particular, we construct INSTRUCTOPENWIKI, a substantial instruction tuning dataset for Open-world IE enriched with a comprehensive corpus, extensive annotations, and diverse instructions.
1 code implementation • 19 Dec 2022 • Xianjun Yang, Kaiqiang Song, Sangwoo Cho, Xiaoyang Wang, Xiaoman Pan, Linda Petzold, Dong Yu
Specifically, zero/few-shot and fine-tuning results show that the model pre-trained on our corpus demonstrates a strong aspect or query-focused generation ability compared with the backbone model.
1 code implementation • 6 Dec 2022 • Pei Chen, Wenlin Yao, Hongming Zhang, Xiaoman Pan, Dian Yu, Dong Yu, Jianshu Chen
However, there has been limited research on the zero-shot KBC settings, where we need to deal with unseen entities and relations that emerge in a constantly growing knowledge base.
no code implementations • 28 Oct 2022 • Xiaoman Pan, Wenlin Yao, Hongming Zhang, Dian Yu, Dong Yu, Jianshu Chen
In this paper, we develop a novel semi-parametric language model architecture, Knowledge-in-Context (KiC), which empowers a parametric text-to-text language model with a knowledge-rich external memory.
Ranked #5 on
Question Answering
on StoryCloze
1 code implementation • 1 Oct 2022 • Zhenhailong Wang, Xiaoman Pan, Dian Yu, Dong Yu, Jianshu Chen, Heng Ji
Notably, our proposed $\text{Zemi}_\text{LARGE}$ outperforms T0-3B by 16% on all seven evaluation tasks while being 3. 9x smaller in model size.
1 code implementation • ACL 2022 • Xiang Yue, Xiaoman Pan, Wenlin Yao, Dian Yu, Dong Yu, Jianshu Chen
And with our pretrained reader, the entire system improves by up to 4% in exact match.
1 code implementation • EMNLP 2021 • Wenlin Yao, Xiaoman Pan, Lifeng Jin, Jianshu Chen, Dian Yu, Dong Yu
We then train a model to identify semantic equivalence between a target word in context and one of its glosses using these aligned inventories, which exhibits strong transfer capability to many WSD tasks.
1 code implementation • NAACL 2021 • Haoyang Wen, Ying Lin, Tuan Lai, Xiaoman Pan, Sha Li, Xudong Lin, Ben Zhou, Manling Li, Haoyu Wang, Hongming Zhang, Xiaodong Yu, Alexander Dong, Zhenhailong Wang, Yi Fung, Piyush Mishra, Qing Lyu, D{\'\i}dac Sur{\'\i}s, Brian Chen, Susan Windisch Brown, Martha Palmer, Chris Callison-Burch, Carl Vondrick, Jiawei Han, Dan Roth, Shih-Fu Chang, Heng Ji
We present a new information extraction system that can automatically construct temporal event graphs from a collection of news documents from multiple sources, multiple languages (English and Spanish for our experiment), and multiple data modalities (speech, text, image and video).
no code implementations • ACL 2020 • Manling Li, Alireza Zareian, Ying Lin, Xiaoman Pan, Spencer Whitehead, Brian Chen, Bo Wu, Heng Ji, Shih-Fu Chang, Clare Voss, Daniel Napierski, Marjorie Freedman
We present the first comprehensive, open source multimedia knowledge extraction system that takes a massive stream of unstructured, heterogeneous multimedia data from various sources and languages as input, and creates a coherent, structured knowledge base, indexing entities, relations, and events, following a rich, fine-grained ontology.
no code implementations • WS 2019 • Xiaoman Pan, Thamme Gowda, Heng Ji, Jonathan May, Scott Miller
Because this multilingual common space directly relates the semantics of contextual words in the source language to that of entities in the target language, we leverage it for unsupervised cross-lingual entity linking.
1 code implementation • WS 2019 • Xiaoman Pan, Kai Sun, Dian Yu, Jianshu Chen, Heng Ji, Claire Cardie, Dong Yu
We focus on multiple-choice question answering (QA) tasks in subject areas such as science, where we require both broad background knowledge and the facts from the given subject-area reference corpus.
1 code implementation • WS 2018 • Qingyun Wang, Xiaoman Pan, Lifu Huang, Boliang Zhang, Zhiying Jiang, Heng Ji, Kevin Knight
We aim to automatically generate natural language descriptions about an input structured knowledge base (KB).
no code implementations • NAACL 2018 • Boliang Zhang, Ying Lin, Xiaoman Pan, Di Lu, Jonathan May, Kevin Knight, Heng Ji
We demonstrate ELISA-EDL, a state-of-the-art re-trainable system to extract entity mentions from low-resource languages, link them to external English knowledge bases, and visualize locations related to disaster topics on a world heatmap.
no code implementations • IJCNLP 2017 • Boliang Zhang, Di Lu, Xiaoman Pan, Ying Lin, Halidanmu Abudukelimu, Heng Ji, Kevin Knight
Current supervised name tagging approaches are inadequate for most low-resource languages due to the lack of annotated data and actionable linguistic knowledge.
no code implementations • ACL 2017 • Xiaoman Pan, Boliang Zhang, Jonathan May, Joel Nothman, Kevin Knight, Heng Ji
The ambitious goal of this work is to develop a cross-lingual name tagging and linking framework for 282 languages that exist in Wikipedia.
no code implementations • COLING 2016 • Dongxu Zhang, Boliang Zhang, Xiaoman Pan, Xiaocheng Feng, Heng Ji, Weiran Xu
Instead of directly relying on word alignment results, this framework combines advantages of rule-based methods and deep learning methods by implementing two steps: First, generates a high-confidence entity annotation set on IL side with strict searching methods; Second, uses this high-confidence set to weakly supervise the model training.
no code implementations • 27 Sep 2016 • Tao Ge, Qing Dou, Xiaoman Pan, Heng Ji, Lei Cui, Baobao Chang, Zhifang Sui, Ming Zhou
We introduce a novel Burst Information Network (BINet) representation that can display the most important information and illustrate the connections among bursty entities, events and keywords in the corpus.
no code implementations • 10 Mar 2016 • Lifu Huang, Jonathan May, Xiaoman Pan, Heng Ji
Recent research has shown great progress on fine-grained entity typing.