no code implementations • NAACL (ACL) 2022 • Wenhao Yu
Retrieval-augmented generation (RAG) methods have been receiving increasing attention from the NLP community and achieved state-of-the-art performance on many NLP downstream tasks.
1 code implementation • EMNLP (ACL) 2021 • Wenhao Yu, Meng Jiang, Zhiting Hu, Qingyun Wang, Heng Ji, Nazneen Rajani
Knowledge-enriched text generation poses unique challenges in modeling and learning, driving active research in several core directions, ranging from integrated modeling of neural representations and symbolic information in the sequential/hierarchical/graphical structures, learning without direct supervisions due to the cost of structured annotation, efficient optimization and inference with massive and global constraints, to language grounding on multiple modalities, and generative reasoning with implicit commonsense knowledge and background knowledge.
no code implementations • ACL 2022 • Chenguang Zhu, Yichong Xu, Xiang Ren, Bill Lin, Meng Jiang, Wenhao Yu
Knowledge in natural language processing (NLP) has been a rising trend especially after the advent of large scale pre-trained models.
no code implementations • 14 Feb 2023 • Zuxin Liu, Zijian Guo, Yihang Yao, Zhepeng Cen, Wenhao Yu, Tingnan Zhang, Ding Zhao
Safe reinforcement learning (RL) trains a constraint satisfaction policy by interacting with the environment.
1 code implementation • 20 Dec 2022 • Pan Lu, Liang Qiu, Wenhao Yu, Sean Welleck, Kai-Wei Chang
Mathematical reasoning is a fundamental aspect of human intelligence and is applicable in various fields, including science, engineering, finance, and everyday life.
no code implementations • 19 Dec 2022 • Soumya Sanyal, Yichong Xu, Shuohang Wang, ZiYi Yang, Reid Pryzant, Wenhao Yu, Chenguang Zhu, Xiang Ren
Logical reasoning of text is an important ability that requires understanding the information present in the text, their interconnections, and then reasoning through them to infer new conclusions.
1 code implementation • 8 Dec 2022 • Wenhao Yu, Chengxiang Zhao, Jiaxin Liu, Yingkai Yang, Xiaohan Ma, Jun Li, Weida Wang, Hong Wang, Ding Zhao
To address these challenges, this paper aims to digitize traffic law comprehensively and provide an application for online monitoring of traffic law violations for autonomous vehicles.
no code implementations • 6 Dec 2022 • Ramij R. Hossain, Tianzhixi Yin, Yan Du, Renke Huang, Jie Tan, Wenhao Yu, YuAn Liu, Qiuhua Huang
We propose a novel model-based-DRL framework where a deep neural network (DNN)-based dynamic surrogate model, instead of a real-world power-grid or physics-based simulation, is utilized with the policy learning framework, making the process faster and sample efficient.
no code implementations • 15 Nov 2022 • Ziniu Hu, Yichong Xu, Wenhao Yu, Shuohang Wang, ZiYi Yang, Chenguang Zhu, Kai-Wei Chang, Yizhou Sun
Answering open-domain questions requires world knowledge about in-context entities.
1 code implementation • 8 Nov 2022 • Liang Peng, Boqi Li, Wenhao Yu, Kai Yang, Wenbo Shao, Hong Wang
Therefore, this paper proposes the "Self-Surveillance and Self-Adaption System" as a systematic approach to online minimize the SOTIF risk, which aims to provide a systematic solution for monitoring, quantification, and mitigation of inherent and external risks.
1 code implementation • 23 Oct 2022 • Wenhao Yu, Chenguang Zhu, Zhihan Zhang, Shuohang Wang, Zhuosheng Zhang, Yuwei Fang, Meng Jiang
However, applying such methods to commonsense reasoning tasks faces two unique challenges, i. e., the lack of a general large-scale corpus for retrieval and a corresponding effective commonsense retriever.
1 code implementation • 12 Oct 2022 • Zhuosheng Zhang, Shuohang Wang, Yichong Xu, Yuwei Fang, Wenhao Yu, Yang Liu, Hai Zhao, Chenguang Zhu, Michael Zeng
Leveraging task-aware annotated data as supervised signals to assist with self-supervised learning on large-scale unlabeled data has become a new trend in pre-training language models.
no code implementations • 7 Oct 2022 • Zhihan Zhang, Wenhao Yu, Chenguang Zhu, Meng Jiang
The entity knowledge is stored in the memory as latent representations, and the memory is pre-trained on Wikipedia along with encoder-decoder parameters.
1 code implementation • 6 Oct 2022 • Mingxuan Ju, Wenhao Yu, Tong Zhao, Chuxu Zhang, Yanfang Ye
In light of this, we propose a novel knowledge Graph enhanced passage reader, namely Grape, to improve the reader performance for open-domain QA.
1 code implementation • 5 Oct 2022 • Mingxuan Ju, Tong Zhao, Qianlong Wen, Wenhao Yu, Neil Shah, Yanfang Ye, Chuxu Zhang
Besides, we observe that learning from multiple philosophies enhances not only the task generalization but also the single task performances, demonstrating that PARETOGNN achieves better task generalization via the disjoint yet complementary knowledge learned from different philosophies.
1 code implementation • 21 Sep 2022 • Wenhao Yu, Dan Iter, Shuohang Wang, Yichong Xu, Mingxuan Ju, Soumya Sanyal, Chenguang Zhu, Michael Zeng, Meng Jiang
We call our method generate-then-read (GenRead), which first prompts a large language model to generate contextutal documents based on a given question, and then reads the generated documents to produce the final answer.
no code implementations • 27 Jul 2022 • Kuang-Huei Lee, Ofir Nachum, Tingnan Zhang, Sergio Guadarrama, Jie Tan, Wenhao Yu
Evolution Strategy (ES) algorithms have shown promising results in training complex robotic control policies due to their massive parallelism capability, simple implementation, effective parameter-space exploration, and fast training time.
no code implementations • 27 Jun 2022 • Yuxiang Yang, Xiangyun Meng, Wenhao Yu, Tingnan Zhang, Jie Tan, Byron Boots
Using only 40 minutes of human demonstration data, our framework learns to adjust the speed and gait of the robot based on perceived terrain semantics, and enables the robot to walk over 6km without failure at close-to-optimal speed.
no code implementations • 7 Apr 2022 • Zhihan Zhang, Wenhao Yu, Mengxia Yu, Zhichun Guo, Meng Jiang
Multi-task learning (MTL) has become increasingly popular in natural language processing (NLP) because it improves the performance of related tasks by exploiting their commonalities and differences.
no code implementations • 28 Mar 2022 • Alejandro Escontrela, Xue Bin Peng, Wenhao Yu, Tingnan Zhang, Atil Iscen, Ken Goldberg, Pieter Abbeel
We also demonstrate that an effective style reward can be learned from a few seconds of motion capture data gathered from a German Shepherd and leads to energy-efficient locomotion strategies with natural gait transitions.
1 code implementation • NAACL (DLG4NLP) 2022 • Wenhao Yu, Chenguang Zhu, Lianhui Qin, Zhihan Zhang, Tong Zhao, Meng Jiang
A set of knowledge experts seek diverse reasoning on KG to encourage various generation outputs.
no code implementations • 5 Mar 2022 • Tsung-Yen Yang, Tingnan Zhang, Linda Luu, Sehoon Ha, Jie Tan, Wenhao Yu
In this paper, we propose a safe reinforcement learning framework that switches between a safe recovery policy that prevents the robot from entering unsafe states, and a learner policy that is optimized to complete the task.
no code implementations • 29 Nov 2021 • Yan Du, Qiuhua Huang, Renke Huang, Tianzhixi Yin, Jie Tan, Wenhao Yu, Xinya Li
Reinforcement learning methods have been developed for the same or similar challenging control problems, but they suffer from training inefficiency and lack of robustness for "corner or unseen" scenarios.
no code implementations • 1 Nov 2021 • Fabio Muratore, Fabio Ramos, Greg Turk, Wenhao Yu, Michael Gienger, Jan Peters
The rise of deep learning has caused a paradigm shift in robotics research, favoring methods that require large amounts of data.
1 code implementation • Findings (ACL) 2022 • Wenhao Yu, Chenguang Zhu, Yuwei Fang, Donghan Yu, Shuohang Wang, Yichong Xu, Michael Zeng, Meng Jiang
In addition to training with the masked language modeling objective, we propose two novel self-supervised pre-training tasks on word and sentence-level alignment between input text sequence and rare word definitions to enhance language modeling representation with dictionary.
no code implementations • ACL 2022 • Donghan Yu, Chenguang Zhu, Yuwei Fang, Wenhao Yu, Shuohang Wang, Yichong Xu, Xiang Ren, Yiming Yang, Michael Zeng
The recent proposed Fusion-in-Decoder (FiD), which is built on top of the pretrained generative model T5, achieves the state-of-the-art performance in the reading module.
no code implementations • 29 Sep 2021 • Michelle Guo, Wenhao Yu, Daniel Ho, Jiajun Wu, Yunfei Bai, Karen Liu, Wenlong Lu
In addition, we perform two studies showing that UC-DiffOSI operates well in environments with changing or unknown dynamics.
1 code implementation • 5 Jun 2021 • Qingkai Zeng, Jinfeng Lin, Wenhao Yu, Jane Cleland-Huang, Meng Jiang
Automatic construction of a taxonomy supports many applications in e-commerce, web search, and question answering.
1 code implementation • NeurIPS 2021 • Tong Zhao, Gang Liu, Daheng Wang, Wenhao Yu, Meng Jiang
However, the causal relationship between the two variables was largely ignored for learning to predict links on a graph.
Ranked #1 on Link Property Prediction on ogbl-ddi
1 code implementation • EMNLP 2021 • Wenhao Yu, Chenguang Zhu, Tong Zhao, Zhichun Guo, Meng Jiang
Generating paragraphs of diverse contents is important in many applications.
no code implementations • 3 Mar 2021 • Yunbo Zhang, Wenhao Yu, C. Karen Liu, Charles C. Kemp, Greg Turk
We produce a final animation by using inverse kinematics to guide a character's arm and hand to match the motion of the manipulation tool such as a knife or a frying pan.
1 code implementation • 16 Feb 2021 • Zhichun Guo, Chuxu Zhang, Wenhao Yu, John Herr, Olaf Wiest, Meng Jiang, Nitesh V. Chawla
The recent success of graph neural networks has significantly boosted molecular property prediction, advancing activities such as drug discovery.
Ranked #1 on Molecular Property Prediction (1-shot)) on Tox21
no code implementations • 8 Feb 2021 • Krzysztof Marcin Choromanski, Deepali Jain, Wenhao Yu, Xingyou Song, Jack Parker-Holder, Tingnan Zhang, Valerii Likhosherstov, Aldo Pacchiano, Anirban Santara, Yunhao Tang, Jie Tan, Adrian Weller
There has recently been significant interest in training reinforcement learning (RL) agents in vision-based environments.
no code implementations • EMNLP (Eval4NLP) 2021 • Qingkai Zeng, Mengxia Yu, Wenhao Yu, Tianwen Jiang, Meng Jiang
It can be used to validate the label consistency (or catches the inconsistency) in multiple sets of NER data annotation.
no code implementations • 13 Jan 2021 • Renke Huang, Yujiao Chen, Tianzhixi Yin, Qiuhua Huang, Jie Tan, Wenhao Yu, Xinya Li, Ang Li, Yan Du
In this paper, we mitigate these limitations by developing a novel deep meta reinforcement learning (DMRL) algorithm.
no code implementations • 11 Dec 2020 • Wenhao Yu, C. Karen Liu, Greg Turk
When used with a set of thresholds, the safety estimator becomes a classifier for switching between the protective policy and the task policy.
no code implementations • 23 Nov 2020 • Zhuo Xu, Wenhao Yu, Alexander Herzog, Wenlong Lu, Chuyuan Fu, Masayoshi Tomizuka, Yunfei Bai, C. Karen Liu, Daniel Ho
General contact-rich manipulation problems are long-standing challenges in robotics due to the difficulty of understanding complicated contact physics.
1 code implementation • 3 Nov 2020 • Ioannis Exarchos, Yifeng Jiang, Wenhao Yu, C. Karen Liu
Transferring reinforcement learning policies trained in physics simulation to the real hardware remains a challenge, known as the "sim-to-real" gap.
no code implementations • Findings of the Association for Computational Linguistics 2020 • Qingkai Zeng, Wenhao Yu, Mengxia Yu, Tianwen Jiang, Tim Weninger, Meng Jiang
The training process of scientific NER models is commonly performed in two steps: i) Pre-training a language model by self-supervised tasks on huge data and ii) fine-tune training with small labelled data.
1 code implementation • 20 Oct 2020 • Tong Zhao, Bo Ni, Wenhao Yu, Zhichun Guo, Neil Shah, Meng Jiang
With Eland, anomaly detection performance at an earlier stage is better than non-augmented methods that need significantly more observed data by up to 15% on the Area under the ROC curve.
1 code implementation • NAACL 2021 • Wenhao Yu, Lingfei Wu, Yu Deng, Qingkai Zeng, Ruchi Mahindru, Sinem Guven, Meng Jiang
In this paper, we propose a novel framework of deep transfer learning to effectively address technical QA across tasks and domains.
3 code implementations • 9 Oct 2020 • Wenhao Yu, Chenguang Zhu, Zaitang Li, Zhiting Hu, Qingyun Wang, Heng Ji, Meng Jiang
To address this issue, researchers have considered incorporating various forms of knowledge beyond the input text into the generation models.
1 code implementation • EMNLP 2020 • Wenhao Yu, Lingfei Wu, Yu Deng, Ruchi Mahindru, Qingkai Zeng, Sinem Guven, Meng Jiang
In recent years, the need for community technical question-answering sites has increased significantly.
2 code implementations • EMNLP 2021 • Xiangyu Dong, Wenhao Yu, Chenguang Zhu, Meng Jiang
Our model has a multi-step decoder that injects the entity types into the process of entity mention generation.
no code implementations • 22 Jun 2020 • Renke Huang, Yujiao Chen, Tianzhixi Yin, Xinya Li, Ang Li, Jie Tan, Wenhao Yu, YuAn Liu, Qiuhua Huang
Load shedding has been one of the most widely used and effective emergency control approaches against voltage instability.
no code implementations • ACL 2020 • Wenhao Yu, Lingfei Wu, Qingkai Zeng, Shu Tao, Yu Deng, Meng Jiang
Existing methods learned semantic representations with dual encoders or dual variational auto-encoders.
no code implementations • WS 2019 • Qingkai Zeng, Mengxia Yu, Wenhao Yu, JinJun Xiong, Yiyu Shi, Meng Jiang
On a scientific concept hierarchy, a parent concept may have a few attributes, each of which has multiple values being a group of child concepts.
1 code implementation • 28 Sep 2019 • Wenhao Yu, Jie Tan, Yunfei Bai, Erwin Coumans, Sehoon Ha
The key idea behind MSO is to expose the same adaptation process, Strategy Optimization (SO), to both the training and testing phases.
no code implementations • 13 May 2019 • Yunbo Zhang, Wenhao Yu, Greg Turk
Our method does this by creating a second reward function that recognizes previously seen state sequences and rewards those by novelty, which is measured using autoencoders that have been trained on state sequences from previously discovered policies.
no code implementations • 4 Mar 2019 • Wenhao Yu, Visak CV Kumar, Greg Turk, C. Karen Liu
We present a new approach for transfer of dynamic robot control policies such as biped locomotion from simulation to real hardware.
1 code implementation • ICLR 2019 • Wenhao Yu, C. Karen Liu, Greg Turk
Transfer learning using domain randomization is a promising approach, but it usually assumes that the target environment is close to the distribution of the training environments, thus relying heavily on accurate system identification.
2 code implementations • 24 Jan 2018 • Wenhao Yu, Greg Turk, C. Karen Liu
Indeed, a standard benchmark for DRL is to automatically create a running controller for a biped character from a simple reward function.
no code implementations • 23 Sep 2017 • Wenhao Yu, C. Karen Liu, Greg Turk
Then, during the specialization training stage we selectively split the weights of the policy based on a per-weight metric that measures the disagreement among the multiple tasks.
1 code implementation • 8 Feb 2017 • Wenhao Yu, Jie Tan, C. Karen Liu, Greg Turk
Together, UP-OSI is a robust control policy that can be used across a wide range of dynamic models, and that is also responsive to sudden changes in the environment.