1 code implementation • ACL 2022 • Dongwon Ryu, Ehsan Shareghi, Meng Fang, Yunqiu Xu, Shirui Pan, Reza Haf
Text-based games (TGs) are exciting testbeds for developing deep reinforcement learning techniques due to their partially observed environments and large action spaces.
no code implementations • 30 Oct 2023 • Mianchu Wang, Rui Yang, Xi Chen, Meng Fang
Offline goal-conditioned RL (GCRL) offers a feasible paradigm to learn general-purpose policies from diverse and multi-task offline datasets.
no code implementations • 23 Oct 2023 • Zihan Zhang, Meng Fang, Ling Chen, Mohammad-Reza Namazi-Rad
In this work, we establish a CIT benchmark consisting of learning and evaluation protocols.
no code implementations • 23 Oct 2023 • Zihan Zhang, Meng Fang, Fanghua Ye, Ling Chen, Mohammad-Reza Namazi-Rad
Dialogue state tracking (DST) plays an important role in task-oriented dialogue systems.
no code implementations • 15 Oct 2023 • Fanghua Ye, Meng Fang, Shenghui Li, Emine Yilmaz
Furthermore, we propose distilling the rewriting capabilities of LLMs into smaller models to reduce rewriting latency.
1 code implementation • 11 Oct 2023 • Zihan Zhang, Meng Fang, Ling Chen, Mohammad-Reza Namazi-Rad, Jun Wang
Although large language models (LLMs) are impressive in solving various tasks, they can quickly be outdated after deployment.
1 code implementation • 20 Sep 2023 • Xin Zheng, Yixin Liu, Zhifeng Bao, Meng Fang, Xia Hu, Alan Wee-Chung Liew, Shirui Pan
Data-centric AI, with its primary focus on the collection, management, and utilization of data to drive AI models and applications, has attracted increasing attention in recent years.
1 code implementation • 29 Aug 2023 • Xinglei Wang, Meng Fang, Zichao Zeng, Tao Cheng
We posit that our research marks a significant paradigm shift in human mobility modelling, transitioning from building complex domain-specific models to harnessing general-purpose LLMs that yield accurate predictions through language instructions.
no code implementations • 29 Jun 2023 • Qiang He, Tianyi Zhou, Meng Fang, Setareh Maghsudi
In ERC, we propose a regularizer that guides the approximation error tending towards the 1-eigensubspace, resulting in a more efficient and stable path of value approximation.
1 code implementation • 25 Jun 2023 • Tianjin Huang, Shiwei Liu, Tianlong Chen, Meng Fang, Li Shen, Vlaod Menkovski, Lu Yin, Yulong Pei, Mykola Pechenizkiy
Despite the fact that adversarial training has become the de facto method for improving the robustness of deep neural networks, it is well-known that vanilla adversarial training suffers from daunting robust overfitting, resulting in unsatisfactory robust generalization.
1 code implementation • 30 May 2023 • Tianjin Huang, Lu Yin, Zhenyu Zhang, Li Shen, Meng Fang, Mykola Pechenizkiy, Zhangyang Wang, Shiwei Liu
We hereby carry out a first-of-its-kind study unveiling that modern large-kernel ConvNets, a compelling competitor to Vision Transformers, are remarkably more effective teachers for small-kernel ConvNets, due to more similar architectures.
1 code implementation • 30 May 2023 • Lu Yin, Gen Li, Meng Fang, Li Shen, Tianjin Huang, Zhangyang Wang, Vlado Menkovski, Xiaolong Ma, Mykola Pechenizkiy, Shiwei Liu
Dynamic sparse training (DST), as a leading sparse training approach, can train deep neural networks at high sparsity from scratch to match the performance of their dense counterparts.
1 code implementation • 28 May 2023 • Yudi Zhang, Yali Du, Biwei Huang, Ziyan Wang, Jun Wang, Meng Fang, Mykola Pechenizkiy
While the majority of current approaches construct the reward redistribution in an uninterpretable manner, we propose to explicitly model the contributions of state and action from a causal perspective, resulting in an interpretable reward redistribution and preserving policy invariance.
1 code implementation • 18 May 2023 • Jiaxu Zhao, Meng Fang, Zijing Shi, Yitong Li, Ling Chen, Mykola Pechenizkiy
We evaluate two popular pretrained Chinese conversational models, CDial-GPT and EVA2. 0, using CHBias.
7 code implementations • 15 May 2023 • Iftitahu Ni'mah, Meng Fang, Vlado Menkovski, Mykola Pechenizkiy
Our proposed framework provides access: (i) for verifying whether automatic metrics are faithful to human preference, regardless of their correlation level to human; and (ii) for inspecting the strengths and limitations of NLG systems via pairwise evaluation.
no code implementations • 6 May 2023 • Dongwon Kelvin Ryu, Meng Fang, Shirui Pan, Gholamreza Haffari, Ehsan Shareghi
Text-based games (TGs) are language-based interactive environments for reinforcement learning.
1 code implementation • 28 Nov 2022 • Tianjin Huang, Tianlong Chen, Meng Fang, Vlado Menkovski, Jiaxu Zhao, Lu Yin, Yulong Pei, Decebal Constantin Mocanu, Zhangyang Wang, Mykola Pechenizkiy, Shiwei Liu
Recent works have impressively demonstrated that there exists a subnetwork in randomly initialized convolutional neural networks (CNNs) that can match the performance of the fully trained dense networks at initialization, without any optimization of the weights of the network (i. e., untrained networks).
no code implementations • 15 Nov 2022 • Qin Zhang, Shangsi Chen, Dongkuan Xu, Qingqing Cao, Xiaojun Chen, Trevor Cohn, Meng Fang
Thus, a trade-off between accuracy, memory consumption and processing speed is pursued.
1 code implementation • 27 Oct 2022 • Yu Cao, Dianqi Li, Meng Fang, Tianyi Zhou, Jun Gao, Yibing Zhan, DaCheng Tao
We present Twin Answer Sentences Attack (TASA), an adversarial attack method for question answering (QA) models that produces fluent and grammatical adversarial contexts while maintaining gold answers.
1 code implementation • 23 Aug 2022 • Lu Yin, Shiwei Liu, Meng Fang, Tianjin Huang, Vlado Menkovski, Mykola Pechenizkiy
We call our method Lottery Pools.
2 code implementations • 16 Jul 2022 • Zhiyin Shao, Xinyu Zhang, Meng Fang, Zhifeng Lin, Jian Wang, Changxing Ding
In PGU, we adopt a set of shared and learnable prototypes as the queries to extract diverse and semantically aligned features for both modalities in the granularity-unified feature space, which further promotes the ReID performance.
no code implementations • 4 Jul 2022 • Jun Rao, Liang Ding, Shuhan Qi, Meng Fang, Yang Liu, Li Shen, DaCheng Tao
Although the vision-and-language pretraining (VLP) equipped cross-modal image-text retrieval (ITR) has achieved remarkable progress in the past two years, it suffers from a major drawback: the ever-increasing size of VLP models restricts its deployment to real-world search scenarios (where the high latency is unacceptable).
no code implementations • 4 Jul 2022 • Yinya Huang, Lemao Liu, Kun Xu, Meng Fang, Liang Lin, Xiaodan Liang
In this work, we propose logic structural-constraint modeling to solve the logical reasoning QA and introduce discourse-aware graph networks (DAGNs).
no code implementations • 30 May 2022 • Lu Yin, Vlado Menkovski, Meng Fang, Tianjin Huang, Yulong Pei, Mykola Pechenizkiy, Decebal Constantin Mocanu, Shiwei Liu
Recent works on sparse neural network training (sparse training) have shown that a compelling trade-off between performance and efficiency can be achieved by training intrinsically sparse neural networks from scratch.
1 code implementation • Findings (NAACL) 2022 • Yibin Lei, Yu Cao, Dianqi Li, Tianyi Zhou, Meng Fang, Mykola Pechenizkiy
Generating high-quality textual adversarial examples is critical for investigating the pitfalls of natural language processing (NLP) models and further promoting their robustness.
1 code implementation • ACL 2022 • Yu Cao, Wei Bi, Meng Fang, Shuming Shi, DaCheng Tao
To alleviate the above data issues, we propose a data manipulation method, which is model-agnostic to be packed with any persona-based dialogue generation model to improve its performance.
1 code implementation • NAACL 2022 • Zihan Zhang, Meng Fang, Ling Chen, Mohammad-Reza Namazi-Rad
Recent work incorporates pre-trained word embeddings such as BERT embeddings into Neural Topic Models (NTMs), generating highly coherent topics.
1 code implementation • ACL 2022 • Yunqiu Xu, Meng Fang, Ling Chen, Yali Du, Joey Tianyi Zhou, Chengqi Zhang
Text-based games provide an interactive way to study natural language processing.
1 code implementation • ICLR 2022 • Rui Yang, Yiming Lu, Wenzhe Li, Hao Sun, Meng Fang, Yali Du, Xiu Li, Lei Han, Chongjie Zhang
In this paper, we revisit the theoretical property of GCSL -- optimizing a lower bound of the goal reaching objective, and extend GCSL as a novel offline goal-conditioned RL algorithm.
no code implementations • 29 Sep 2021 • Meng Fang, Yunqiu Xu, Yali Du, Ling Chen, Chengqi Zhang
In a variety of text-based games, we show that this simple method results in competitive performance for agents.
no code implementations • 29 Sep 2021 • Zhihao Cheng, Li Shen, Meng Fang, Liu Liu, DaCheng Tao
Imitation Learning (IL) merely concentrates on reproducing expert behaviors and could take dangerous actions, which is unbearable in safety-critical scenarios.
1 code implementation • Findings (EMNLP) 2021 • Yunqiu Xu, Meng Fang, Ling Chen, Yali Du, Chengqi Zhang
Deep reinforcement learning provides a promising approach for text-based games in studying natural language communication between humans and artificial agents.
Hierarchical Reinforcement Learning
reinforcement-learning
+2
1 code implementation • Findings (EMNLP) 2021 • Iftitahu Ni'mah, Meng Fang, Vlado Menkovski, Mykola Pechenizkiy
The ability to detect Out-of-Domain (OOD) inputs has been a critical requirement in many real-world NLP applications.
no code implementations • 9 Jul 2021 • Hao Sun, Ziping Xu, Meng Fang, Zhenghao Peng, Jiadong Guo, Bo Dai, Bolei Zhou
Safe exploration is crucial for the real-world application of reinforcement learning (RL).
no code implementations • 1 Jul 2021 • Rui Yang, Meng Fang, Lei Han, Yali Du, Feng Luo, Xiu Li
Replacing original goals with virtual goals generated from interaction with a trained dynamics model leads to a novel relabeling method, model-based relabeling (MBR).
2 code implementations • NAACL 2021 • Yinya Huang, Meng Fang, Yu Cao, LiWei Wang, Xiaodan Liang
The model encodes discourse information as a graph with elementary discourse units (EDUs) and discourse relations, and learns the discourse-aware features via a graph network for downstream QA tasks.
Ranked #24 on
Reading Comprehension
on ReClor
1 code implementation • 2 Mar 2021 • Yu Cao, Liang Ding, Zhiliang Tian, Meng Fang
Dialogue generation models face the challenge of producing generic and repetitive responses.
no code implementations • 1 Jan 2021 • Hao Sun, Ziping Xu, Meng Fang, Yuhang Song, Jiechao Xiong, Bo Dai, Zhengyou Zhang, Bolei Zhou
Despite the remarkable progress made by the policy gradient algorithms in reinforcement learning (RL), sub-optimal policies usually result from the local exploration property of the policy gradient update.
no code implementations • 1 Jan 2021 • Yali Du, Yifan Zhao, Meng Fang, Jun Wang, Gangyan Xu, Haifeng Zhang
Dealing with multi-agent control in networked systems is one of the biggest challenges in Reinforcement Learning (RL) and limited success has been presented compared to recent deep reinforcement learning in single-agent domain.
no code implementations • 24 Dec 2020 • Yinya Huang, Meng Fang, Xunlin Zhan, Qingxing Cao, Xiaodan Liang, Liang Lin
It is crucial since the quality of the evidence is the key to answering commonsense questions, and even determines the upper bound on the QA systems performance.
1 code implementation • 27 Nov 2020 • Lei Han, Jiechao Xiong, Peng Sun, Xinghai Sun, Meng Fang, Qingwei Guo, Qiaobo Chen, Tengfei Shi, Hongsheng Yu, Xipeng Wu, Zhengyou Zhang
We show that with orders of less computation scale, a faithful reimplementation of AlphaStar's methods can not succeed and the proposed techniques are necessary to ensure TStarBot-X's competitive performance.
1 code implementation • 25 Nov 2020 • Peng Sun, Jiechao Xiong, Lei Han, Xinghai Sun, Shuxing Li, Jiawei Xu, Meng Fang, Zhengyou Zhang
This poses non-trivial difficulties for researchers or engineers and prevents the application of MARL to a broader range of real-world problems.
1 code implementation • NeurIPS 2020 • Yunqiu Xu, Meng Fang, Ling Chen, Yali Du, Joey Tianyi Zhou, Chengqi Zhang
We study reinforcement learning (RL) for text-based games, which are interactive simulations in the context of natural language.
no code implementations • 16 Oct 2020 • Zhihao Cheng, Liu Liu, Aishan Liu, Hao Sun, Meng Fang, DaCheng Tao
By contrast, this paper proves that LfO is almost equivalent to LfD in the deterministic robot environment, and more generally even in the robot environment with bounded randomness.
1 code implementation • Findings of the Association for Computational Linguistics 2020 • Yu Cao, Wei Bi, Meng Fang, DaCheng Tao
In this work, we study dialogue models with multiple input sources adapted from the pretrained language model GPT2.
no code implementations • 11 Jun 2020 • Hao Sun, Ziping Xu, Yuhang Song, Meng Fang, Jiechao Xiong, Bo Dai, Bolei Zhou
However, PG algorithms rely on exploiting the value function being learned with the first-order update locally, which results in limited sample efficiency.
1 code implementation • NeurIPS 2019 • Yali Du, Lei Han, Meng Fang, Ji Liu, Tianhong Dai, DaCheng Tao
A great challenge in cooperative decentralized multi-agent reinforcement learning (MARL) is generating diversified behaviors for each individual agent when receiving only a team reward.
Multi-agent Reinforcement Learning
reinforcement-learning
+3
1 code implementation • NeurIPS 2019 • Meng Fang, Tianyi Zhou, Yali Du, Lei Han, Zhengyou Zhang
This ``Goal-and-Curiosity-driven Curriculum Learning'' leads to ``Curriculum-guided HER (CHER)'', which adaptively and dynamically controls the exploration-exploitation trade-off during the learning process via hindsight experience selection.
1 code implementation • 13 Nov 2019 • Yu Cao, Meng Fang, Baosheng Yu, Joey Tianyi Zhou
On the other hand, it further reduces domain distribution discrepancy through conditional adversarial learning across domains.
1 code implementation • 10 Sep 2019 • Yutai Hou, Meng Fang, Wanxiang Che, Ting Liu
The framework builds a user simulator by first generating diverse dialogue data from templates and then build a new State2Seq user simulator on the data.
1 code implementation • 26 Jul 2019 • Tingguang Li, Weitao Xi, Meng Fang, Jia Xu, Max Qing-Hu Meng
We present a learning-based approach to solving a Rubik's cube with a multi-fingered dexterous hand.
Robotics
2 code implementations • 20 Jul 2019 • Qing Wang, Jiechao Xiong, Lei Han, Meng Fang, Xinghai Sun, Zhuobin Zheng, Peng Sun, Zhengyou Zhang
We introduce Arena, a toolkit for multi-agent reinforcement learning (MARL) research.
no code implementations • 6 Jul 2019 • Xiaomeng Li, Lequan Yu, Chi-Wing Fu, Meng Fang, Pheng-Ann Heng
However, the importance of feature embedding, i. e., exploring the relationship among training samples, is neglected.
no code implementations • ACL 2019 • Joey Tianyi Zhou, Hao Zhang, Di Jin, Hongyuan Zhu, Meng Fang, Rick Siow Mong Goh, Kenneth Kwok
We propose a new neural transfer method termed Dual Adversarial Transfer Network (DATNet) for addressing low-resource Named Entity Recognition (NER).
1 code implementation • ICLR 2019 • Meng Fang, Cheng Zhou, Bei Shi, Boqing Gong, Jia Xu, Tong Zhang
Dealing with sparse rewards is one of the most important challenges in reinforcement learning (RL), especially when a goal is dynamic (e. g., to grasp a moving object).
1 code implementation • NAACL 2019 • Yu Cao, Meng Fang, DaCheng Tao
Graph convolutional networks are used to obtain a relation-aware representation of nodes for entity graphs built from documents with multi-level features.
1 code implementation • 9 Sep 2018 • Yali Du, Meng Fang, Jin-Feng Yi, Jun Cheng, DaCheng Tao
First, we initialize an adversarial example with a gray color image on which every pixel has roughly the same importance for the target model.
1 code implementation • 2 May 2018 • Songyou Peng, Le Zhang, Yutong Ban, Meng Fang, Stefan Winkler
In this paper, we comprehensively describe the methodology of our submissions to the One-Minute Gradual-Emotion Behavior Challenge 2018.
1 code implementation • EMNLP 2017 • Meng Fang, Yuan Li, Trevor Cohn
Active learning aims to select a small subset of data for annotation such that a classifier learned on the data is highly accurate.
1 code implementation • ACL 2017 • Meng Fang, Trevor Cohn
Cross-lingual model transfer is a compelling and popular method for predicting annotations in a low-resource language, whereby parallel corpora provide a bridge to a high-resource language and its associated annotated corpora.
no code implementations • 19 Aug 2016 • Yang Wang, Wenjie Zhang, Lin Wu, Xuemin Lin, Meng Fang, Shirui Pan
Multi-view spectral clustering, which aims at yielding an agreement or consensus data objects grouping across multi-views with their graph laplacian matrices, is a fundamental clustering problem.
no code implementations • CONLL 2016 • Meng Fang, Trevor Cohn
Cross lingual projection of linguistic annotation suffers from many sources of bias and noise, leading to unreliable annotations that cannot be used directly.
no code implementations • CVPR 2015 • Chen Gong, DaCheng Tao, Wei Liu, Stephen J. Maybank, Meng Fang, Keren Fu, Jie Yang
In the teaching-to-learn step, a teacher is designed to arrange the regions from simple to difficult and then assign the simplest regions to the learner.
no code implementations • 11 Mar 2014 • Meng Fang, Jie Yin, Xingquan Zhu
In this paper, we propose a new transfer learning algorithm that attempts to transfer common latent structure features across the source and target networks.