Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileAcademic lineage
Possible advisorsa guess from early papers, not confirmed
- Suggested from co-authorship
Is this you? Claim this profile to confirm or dismiss it.
Works126 from public data
- Check Your Facts and Try Again: Improving Large Language Models with External Knowledge and Automated Feedback552
A LLM-Augmenter system, which augments a black-box LLM with a set of plug-and-play modules to significantly reduces ChatGPT's hallucinations without sacrificing the fluency and informativeness of its responses.
- An Introduction to Computational Networks and the Computational Network Toolkit403
The computational network toolkit (CNTK), an implementation of CN that supports both GPU and CPU, is introduced and the architecture and the key components of the CNTK are described, the command line options to use C NTK, and the network definition and model editing language are described.
- Latent Action Pretraining from Videos382
Experimental results demonstrate that the LAPA method significantly outperforms existing techniques that train robot manipulation policies from large-scale videos and outperforms the state-of-the-art VLA model trained with robotic action labels on real-world manipulation tasks that require language conditioning, generalization to unseen objects, and semantic generalization to unseen instructions.
- 317
- Task-oriented Dialogue System for Automatic Diagnosis255
Experimental results on this dialogue system show that additional symptoms extracted from conversation can greatly improve the accuracy for disease identification and the dialogue system is able to collect these symptoms automatically and make a better diagnosis.
- S oloist : BuildingTask Bots at Scale with Transfer Learning and Machine Teaching230
A new method that uses transfer learning and machine teaching to build task bots at scale, Soloist, is presented, which parameterize classical modular task-oriented dialog systems using a Transformer-based auto-regressive language model, which subsumes different dialog modules into a single neural model.
- Few-shot Natural Language Generation for Task-Oriented Dialog213
FewshotWOZ is presented, the first NLG benchmark to simulate the few-shot learning setting in task-oriented dialog systems, and the proposed SC-GPT model significantly outperforms existing methods, measured by various automatic metrics and human evaluations.
- Optimus: Organizing Sentences via Pre-trained Modeling of a Latent Space205
This paper proposes the first large-scale language VAE model, Optimus, a universal latent embedding space for sentences that is first pre-trained on large text corpus, and then fine-tuned for various language generation and understanding tasks.
- Composite Task-Completion Dialogue Policy Learning via Hierarchical Deep Reinforcement Learning197
This paper addresses the travel planning task by formulating the task in the mathematical framework of options over Markov Decision Processes (MDPs), and proposing a hierarchical deep reinforcement learning approach to learning a dialogue manager that operates at different temporal scales.
- Magma: A Foundation Model for Multimodal AI Agents185
Magma is a foundation model that serves multimodal AI agentic tasks in both the digital and physical worlds and creates new state-of-the-art results on UI navigation and robotic manipulation tasks, outperforming previous models that are specifically tailored to these tasks.
- Deep Dyna-Q: Integrating Planning for Task-Completion Dialogue Policy Learning184
Deep Dyna-Q is presented, which to the authors' knowledge is the first deep RL framework that integrates planning for task-completion dialogue policy learning and incorporates into the dialogue agent a model of the environment, referred to as the world model, to mimic real user response and generate simulated experience.
- Explanations from Large Language Models Make Small Reasoners Better181
This paper considers the problem of leveraging the explanations generated by LLM to improve the training of small reasoners, which are more favorable in real-production deployment due to their low cost.
- 175
- Integrating planning for task-completion dialogue policy learning.147
This paper addresses challenges of training a task-completion dialogue agent with real users via reinforcement learning by integrating planning into the dialogue policy learning based on Dyna-Q framework, and provides a more sample-efficient approach to learn the dialogue polices.
- CollabLLM: From Passive Responders to Active Collaborators145
CollabLLM is introduced, a novel and general training framework that enhances multiturn human-LLM collaboration and estimates the long-term contribution of responses using Multiturn-aware Rewards by reinforcement fine-tuning these rewards.
- 136
- SOLOIST: Few-shot Task-Oriented Dialog with A Single Pre-trained Auto-regressive Model127
A new method SOLOIST is presented, which uses transfer learning to efficiently build task-oriented dialog systems at scale using a Transformer-based auto-regressive language model, which subsumes different dialog modules into a single neural model.
- ConvLab-2: An Open-Source Toolkit for Building, Evaluating, and Diagnosing Dialogue Systems121
ConvLab-2, an open-source toolkit that enables researchers to build task-oriented dialogue systems with state-of-the-art models, perform an end-to-end evaluation, and diagnose the weakness of systems is presented.
- 120
- ConvLab: Multi-Domain End-to-End Dialog System Platform117
ConvLab is presented, an open-source multi-domain end-to-end dialog system platform that enables researchers to quickly set up experiments with reusable components and compare a large set of different approaches, ranging from conventional pipeline systems to end- to-end neural models, in common environments.
- Few-Shot Named Entity Recognition: An Empirical Baseline Study110
Three orthogonal schemes to improve model generalization ability in few-shot settings are investigated: meta-learning to construct prototypes for different entity types, task-specific supervised pre-training on noisy web data to extract entity-related representations and self-training to leverage unlabeled in-domain data.
- Evaluating the Instruction-Following Robustness of Large Language Models to Prompt Injection92
A benchmark is established to evaluate the robustness of instruction-following LLMs against prompt injection attacks, assessing their ability to discern which instructions to follow and which to disregard, and shows that certain models are excessively inclined to prioritize embedded instructions in prompts.
- 89
- GODEL: Large-Scale Pre-Training for Goal-Directed Dialog80
GODEL (Grounded Open Dialogue Language Model) is introduced, a large pre-trained language model for dialog that leverages a new phase of grounded pre-training designed to better support adapting GODEL to a wide range of downstream dialog tasks that require information external to the current conversation to produce good responses.
- Adversarial Advantage Actor-Critic Model for Task-Completion Dialogue Policy Learning76
This paper presents a new method - adversarial advantage actor-critic (Adversarial A2C), which significantly improves the efficiency of dialogue policy learning in task-completion dialogue systems.
- Overview of the Ninth Dialog System Technology Challenge: DSTC973
The task definition is described, provided datasets, baselines, and evaluation setup for each track, and the results of the submitted systems are summarized to highlight the general trends of the state-of-the-art technologies for the tasks.
- ExACT: Teaching AI Agents to Explore with Reflective-MCTS and Exploratory Learning67
ExACT, an approach to combine test-time search and self-learning to build o1-like models for agentic applications, is presented, suggesting a promising research direction to enhance VLMs' capabilities for agentic applications via test-time search and self-learning.
- Knowledge-Grounded Dialogue Generation with a Unified Knowledge Representation63
PLUG is a language model that homogenizes different knowledge sources to a unified knowledge representation for knowledge-grounded dialogue generation tasks and achieves comparable performance with state-of-the-art methods in the fully-supervised setting and significantly outperforms other approaches in zero-shot and few-shot settings.
- 63
- Recurrent Neural Networks with External Memory for Spoken Language Understanding61
This work proposes to use an external memory to improve memorisation capability of RNNs and demonstrates that the proposed model was able to achieve the state-of-the-art results.
- ThetaEvolve: Test-time Learning on Open Problems58
ThetaEvolve is the first evolving framework that enable a small open-source model, like DeepSeek-R1-0528-Qwen3-8B, to achieve new best-known bounds on open problems (circle packing and first auto-correlation inequality) mentioned in AlphaEvolve.
- Task-oriented Dialogue System for Automatic Disease Diagnosis via Hierarchical Reinforcement Learning.58
Experimental results on both self-constructed real-world and synthetic datasets demonstrate that the proposed hierarchical framework achieves higher accuracy in disease diagnosis compared with existing systems.
- ValueNet: A New Dataset for Human Value Driven Dialogue System57
This work presents a new large-scale human value dataset called ValueNet, which contains human attitudes on 21,374 text scenarios and is the first one trying to incorporate a value model into emotionally intelligent dialogue systems.
- Is Your Goal-Oriented Dialog Model Performing Really Well? Empirical Analysis of System-wise Evaluation54
This paper performs a system-wise evaluation and presents an empirical analysis on different types of dialog systems which are composed of different modules in different settings to show that a pipeline dialog system trained using fine-grained supervision signals at different component levels often obtains better performance.
- RADDLE: An Evaluation Benchmark and Analysis Platform for Robust Task-oriented Dialog Systems53
The RADDLE benchmark is introduced, a collection of corpora and tools for evaluating the performance of models across a diverse set of domains designed to favor and encourage models with a strong generalization ability and evaluate existing models less than satisfactory in robustness evaluation.
- 52
- 50
- Word Embedding Based Correlation Model for Question/Answer Matching49
A Word Embedding based Correlation (WEC) model is proposed by integrating advantages of both the translation model and word embedding, which can score their co-occurrence probability in Q&A pairs and leverage the continuity and smoothness of continuous space word representation.
- 47
- Phi-4-Mini-Reasoning: Exploring the Limits of Small Reasoning Language Models in Math43
This work presents a systematic training recipe for SLMs that consists of large-scale mid-training on diverse distilled long-CoT data, supervised fine-tuning on high-quality long-CoT data, Rollout DPO leveraging a carefully curated preference dataset, and Reinforcement Learning (RL) with Verifiable Reward.
- 36
- 35
- Hierarchical reinforcement learning for automatic disease diagnosis32
Inspired by the offline consultation process, this work proposes to integrate a hierarchical policy structure of two levels into the dialogue system for policy learning that achieves higher accuracy and symptom recall in disease diagnosis compared with existing systems.
- Collaborative personal profiling for web service ranking and recommendation32
Inspired by the idea that the service composition pattern is significant hint for service selection, a personal profiling mechanism is proposed to improve ranking and recommendation performance.
- 31
- 29
- Guided Dialogue Policy Learning without Adversarial Learning in the Loop28
The proposed decompose the adversarial training into two steps, which achieves a remarkable task success rate using both on-policy and off-policy reinforcement learning methods and has potential to transfer knowledge from existing domains to a new domain.
- An empirical analysis of different sparse penalties for autoencoder in unsupervised feature learning28
- 27
- 27
- Data Augmentation for Spoken Language Understanding via Pretrained Models26
A data augmentation method with pretrained language models to boost the variability and accuracy of generated utterances and investigate and propose solutions to two previously overlooked scenarios of data scarcity in SLU.
- Results of the Multi-Domain Task-Completion Dialog Challenge26
An overview of the “Multi-domain Task Completion” track (Track 1) at the 8th Dialog System Technology Challenge (DSTC-8) is provided, seeking to develop models that predict user responses when only limited in-domain data is available.
- Auto-encoder based bagging architecture for sentiment analysis26
An auto-encoder-based bagging architecture is proposed and experiment on real-world datasets reveals its potential and it is believed that this method can offer the researchers in the community further insight into bagging oriented solution for sentimental analysis.
- Structural information aware deep semi-supervised recurrent neural network for sentiment analysis26
A novel sentiment analysis model is proposed based on recurrent neural network, which takes the partial document as input and then the next parts to predict the sentiment label distribution rather than the next word.
- Learning Efficient Dialogue Policy from Demonstrations through Shaping25
An imitation model is used to distill knowledge from demonstrations, based on which policy shaping estimates feedback on how the agent should act in policy space is estimated, and reward shaping is incorporated to bonus state-actions similar to demonstrations explicitly in value space encouraging better exploration.
- 22
- SocAoG: Incremental Graph Parsing for Social Relation Inference in Dialogues21
This work model the social network as an And-or Graph, named SocAoG, for the consistency of relations among a group and leveraging attributes as inference cues, and proposes an incremental structure prediction task.
- Conversation Learner - A Machine Teaching Tool for Building Dialog Managers for Task-Oriented Dialog Systems21
Conversation Learner is presented, a machine teaching tool for building dialog managers that combines the best of both approaches by enabling dialog authors to create a dialog flow using familiar tools, converting the dialog flow into a parametric model (e.g., neural networks), and allowingDialog authors to improve the dialog manager over time by leveraging user-system dialog logs as training data through a machine learning interface.
- 20
- Engage the Public: Poll Question Generation for Social Media Posts20
The results show that the model outperforms the popular S2S models without exploiting topics from comments and the dual decoder design can further benefit the prediction of both questions and answers.
- Using Content-level Structures for Summarizing Microblog Repost Trees19
A novel repost tree summarization framework is proposed by effectively differentiating two kinds of messages on repost trees called leaders and followers, which are derived from contentlevel structure information, i.e., contents of messages and the reposting relations.
- 16
- Data Augmentation for Spoken Language Understanding via Pretrained Language Models16
A data augmentation method using pretrained language models to boost the variability and accuracy of generated utterances and investigate and propose solutions to two previously overlooked semi-supervised learning scenarios of data scarcity in SLU.
- 16
- 15
- Trending Sentiment-Topic Detection on Twitter15
This work is based on a non-parametric supervised real-time trending topic detection model with sentimental feature that demonstrates impressive effectiveness with 82.3% recall and surpasses all the competitors.
- Composite Task-Completion Dialogue System via Hierarchical Deep Reinforcement Learning.14
This paper addresses the travel planning task by formulating the task in the mathematical framework of options over Markov Decision Processes (MDPs), and proposing a hierarchical deep reinforcement learning approach to learning a dialogue manager that operates at different temporal scales.
- Dyna-Think: Synergizing Reasoning, Acting, and World Model Simulation in AI Agents13
Dyna-Think, a thinking framework that integrates planning with an internal world model with reasoning and acting to enhance AI agent performance, is proposed and a promising research direction to integrate world model simulation into AI agents to enhance their reasoning, planning, and acting capabilities is suggested.
- 13
- GUI-Libra: Training Native GUI Agents to Reason and Act with Action-aware Supervision and Partially Verifiable RL12
This work presents GUI-Libra, a tailored training recipe that consistently improves both step-wise accuracy and end-to-end task completion and introduces success-adaptive scaling to downweight unreliable negative gradients.
- SynthAgent: Adapting Web Agents with Synthetic Supervision12
This paper proposes SynthAgent, a fully synthetic supervision framework that aims at improving synthetic data quality via dual refinement of both tasks and trajectories, and demonstrates that SynthAgent outperforms existing synthetic data methods, validating the importance of high-quality synthetic supervision.
- Toward Self-Learning End-to-End Task-oriented Dialog Systems12
SL-Agent is proposed, a novel self-learning framework for building end-to-end task bots that enables task bots to automatically adapt to changing environments by learning from the unlabeled human-bot dialog logs accumulated after deployment via reinforcement learning with the incorporated reward model.
- OPERA: Harmonizing Task-Oriented Dialogs and Information Seeking Experience11
A unified model OPERA (Open-book End-to-end Task-oriented Dialog) which can appropriately access explicit and implicit external knowledge to tackle the OB-TOD task is proposed and Experimental results show that OPERA outperforms closed-book baselines, highlighting the value of both types of knowledge.
- 10
- Fine-Grained Self-Endorsement Improves Factuality and Reasoning9
This work proposes a self-endorsement framework that leverages the fine-grained fact-level comparisons across multiple sampled responses that can better alleviate hallucinations, especially for longform generation tasks.
- ArK: Augmented Reality with Knowledge Interactive Emergent Ability9
It is shown that the ArK approach, combined with large foundation models, significantly improves the quality of generated 2D/3D scenes, compared to baselines, demonstrating the potential benefit of incorporating ArK in generative AI for applications such as metaverse and gaming simulation.
- 9
- 9
- Orchard: An Open-Source Agentic Modeling Framework8
Orchard Env, a lightweight Kubernetes-native environment service that provides reusable primitives for sandbox lifecycle management across task domains, agent harnesses, and training stages, and three agentic modeling recipes are presented, demonstrating that a lightweight, open, harness-agnostic environment layer enables reusable agentic data, training recipes, and evaluation protocols across domains.
- Dyna-Mind: Learning to Simulate from Experience for Better AI Agents8
Dyna-Mind, a two-stage training framework that explicitly teaches (V)LM agents to integrate simulation ability into AI agents, and Dyna-GRPO, an online reinforcement learning method that leverages outcome and interaction-level signals to learn better policies for long-horizon, planning-intensive tasks are introduced.
- 8
- 7
- 6
- Building Personalized Simulator for Interactive Search6
A simulator is proposed to employ to mimic the environment for the offline training of the agent and modelbased approach is used to train the simulator and is able to use the data efficiently.
- ACE: Automatic Colloquialism, Typographical and Orthographic Errors Detection for Chinese Language6
A system called ACE for Automatic Colloquialism and Errors detection for written Chinese, based on the combination of N-gram model and rule-base model, which can be extended to detect other dialects.
- The Tool Illusion: Rethinking Tool Use in Web Agents5
This study revisits tool use in web agents through an extensive and carefully controlled study across diverse tool sources, backbone models, tool-use frameworks, and evaluation benchmarks, and hopes it provides a more reliable empirical basis and inspires future research on tool-use web agents.
- OpenWebRL: Demystifying Online Multi-turn Reinforcement Learning for Visual Web Agents5
This work introduces OpenWebRL, an open framework for training visual web agents with online multi-turn RL on real websites, and trains OpenWebRL-4B, which establishes a new open-source state of the art on challenging live-web benchmarks.
- Latent Recurrent Transformer: Architecture Exploration, Training Strategies, and Scaling Behavior5
Latent Recurrent Transformer is studied, a lightweight augmentation of autoregressive transformers that reuses a high-level source-layer hidden state from the previous token as recurrent memory for the next token, while retaining one-forward-per-token decoding with 9% latency overhead over the standard Transformer.
- Multi-domain Task-oriented Dialog Challenge II5
This proposal extends the tasks by incorporating new datasets, creating new sub-tasks, and providing a new development platform, and specifically focuses on two aspects of dialog systems: language portability and end-to-end system complexity.
- 4
- 4
- 4
- 4
- 4
- 4
- 3
- 2
- 2
- 2
- 2
- 1
- 1
- 1
- 1
- 1
- 1
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- May I take your order? A Neural Model for Extracting Structured Information from Conversations–
This paper develops a sequence-to-sequence model that is able to map from unstructured conversational input to the structured form that is conveyed to the kitchen and appears on the customer receipt.
- –
- 基于情感时间序列的微博热点主题检测–
A non-parametric supervised real-time-trending topic-detection model with a sentimental feature that detects trending sentimental topic in a short time and demonstrates impressive effectiveness with an 82.3% recall rate, surpassing all of the competitors.
- –
Publication data from OpenAlex, with missing venues and authors filled in from Crossref; citation counts are the higher of OpenAlex and Semantic Scholar, last synced 2026-10-16. One-sentence summaries under some papers are written by Semantic Scholar’s model. Citation counts may be lower than on Google Scholar, which indexes more sources.
Report an error
Wrong papers, two people merged into one, or a profile that should not be here? Tell us and we will fix or hide it.