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
- Shouxun LinSuggested from co-authorship
Is this you? Claim this profile to confirm or dismiss it.
Works196 from public data
- Modeling Coverage for Neural Machine Translation786
This paper proposes coverage-based NMT, which maintains a coverage vector to keep track of the attention history and improves both translation quality and alignment quality over standard attention- based NMT.
- Do NOT Think That Much for 2+3=? On the Overthinking of o1-Like LLMs586
This paper introduces novel efficiency metrics from both outcome and process perspectives to evaluate the rational use of computational resources by o1-like models, and proposes strategies to mitigate overthinking, streamlining reasoning processes without compromising accuracy.
- On the "naturalness" of buggy code305
It is found that code with bugs tends to be more entropic (i.e. unnatural), becoming less so as bugs are fixed, suggesting that entropy may be a valid, simple way to complement the effectiveness of PMD or FindBugs, and that search-based bug-fixing methods may benefit from using entropy both for fault-localization and searching for fixes.
- 304
- On the localness of software289
A novel cache language model that consists of both an n-gram and an added “ cache" component to exploit localness, which finds that human-written programs are localized: they have useful local regularities that can be captured and exploited.
- DeepMath-103K: A Large-Scale, Challenging, Decontaminated, and Verifiable Mathematical Dataset for Advancing Reasoning282
DeepMath-103K is introduced, a large-scale mathematical dataset designed with high difficulty, rigorous decontamination against numerous benchmarks, and verifiable answers for rule-based RL reward that fosters the development of generalizable and advancing reasoning.
- Document-Level Machine Translation with Large Language Models219
An in-depth evaluation of LLMs' ability on discourse modeling using document-level machine translation (MT) as a testbed finds that LLMs have demonstrated superior performance and show potential to become a new paradigm for document- level translation.
- Neural Machine Translation with Reconstruction212
Experiments show that the proposed framework significantly improves the adequacy of NMT output and achieves superior translation result over state-of-the-art NMT and statistical MT systems.
- Exploiting Cross-Sentence Context for Neural Machine Translation209
This paper proposes a cross-sentence context-aware approach and investigates the influence of historical contextual information on the performance of neural machine translation (NMT).
- Modeling Localness for Self-Attention Networks196
This work cast localness modeling as a learnable Gaussian bias, which indicates the central and scope of the local region to be paid more attention in self-attention networks, to maintain the strength of capturing long distance dependencies while enhance the ability of capturing short-range dependencies.
- Multi-Head Attention with Disagreement Regularization190
This work proposes three types of disagreement regularization, which respectively encourage the subspace, the attended positions, and the output representation associated with each attention head to be different from other heads.
- Learning to Remember Translation History with a Continuous Cache189
This work proposes to augment NMT models with a very light-weight cache-like memory network, which stores recent hidden representations as translation history and the probability distribution over generated words is updated online depending on the translation history retrieved from the memory.
- Thoughts Are All Over the Place: On the Underthinking of o1-Like LLMs172
A decoding strategy with thought switching penalty TIP that discourages premature transitions between thoughts, encouraging deeper exploration of each reasoning path and contributes to understanding reasoning inefficiencies in o1-like LLMs and offers a practical solution to enhance their problem-solving capabilities.
- Towards Robust Neural Machine Translation169
Experimental results on Chinese-English, English-German and English-French translation tasks show that the proposed approaches can not only achieve significant improvements over strong NMT systems but also improve the robustness of NMT models.
- Context Gates for Neural Machine Translation152
Context gates are proposed which dynamically control the ratios at which source and target contexts contribute to the generation of target words and can enhance both the adequacy and fluency of NMT with more careful control of the information flow from contexts.
- EmpDG: Multi-resolution Interactive Empathetic Dialogue Generation150
A multi-resolution adversarial model – EmpDG is proposed, to generate more empathetic responses and an interactive adversarial learning framework which exploits the user feedback, to identify whether the generated responses evoke emotion perceptivity in dialogues.
- Modeling Source Syntax for Neural Machine Translation150
This paper shows that source syntax can be explicitly incorporated into NMT effectively to provide further improvements, and linearize parse trees of source sentences to obtain structural label sequences.
- Convolutional Self-Attention Networks137
Novel convolutional self-attention networks are proposed, which offer SANs the abilities to strengthen dependencies among neighboring elements, and model the interaction between features extracted by multiple attention heads.
- On the diversity of multi-head attention134
This work introduces a disagreement regularization to explicitly encourage the diversity among multiple attention heads, and proposes to better capture the diverse information distributed in the extracted partial-representations with the routing-by-agreement algorithm.
- Context-Aware Self-Attention Networks124
This work proposes to contextualize the transformations of the query and key layers, which are used to calculates the relevance between elements, and leverage the internal representations that embed both global and deep contexts, thus avoid relying on external resources.
- Exploring Human-Like Translation Strategy with Large Language Models118
This work proposes the MAPS framework, which stands for Multi-Aspect Prompting and Selection, and enables LLMs first to analyze the given source sentence and induce three aspects of translation-related knowledge to guide the final translation process.
- Skeleton-to-Response: Dialogue Generation Guided by Retrieval Memory99
This paper proposes a new framework which exploits retrieval results via a skeleton-to-response paradigm and shows that this approach significantly improves the informativeness of the generated responses.
- Exploiting Deep Representations for Neural Machine Translation98
This work proposes to simultaneously expose all of the top layers of encoder and decoder with layer aggregation and multi-layer attention mechanisms, and introduces an auxiliary regularization term to encourage different layers to capture diverse information.
- Neural Machine Translation Advised by Statistical Machine Translation95
Experimental results on Chinese-English translation show that the proposed approach achieves significant and consistent improvements over state-of-the-art NMT and SMT systems on multiple NIST test sets.
- On the Inference Calibration of Neural Machine Translation94
This work provides in-depth analyses of the correlation between calibration and translation performance as well as linguistic properties of miscalibration and reports a number of interesting findings that might help humans better analyze, understand and improve NMT models.
- Retrieval-guided Dialogue Response Generation via a Matching-to-Generation Framework91
A novel framework in which the skeleton extraction is made by an interpretable matching model and the following skeleton-guided response generation is accomplished by a separately trained generator is presented.
- Modeling Recurrence for Transformer87
This work proposes to directly model recurrence for Transformer with an additional recurrence encoder, and introduces a novel attentive recurrent network to leverage the strengths of both attention models and recurrent networks.
- GliDe with a CaPE: A Low-Hassle Method to Accelerate Speculative Decoding84
GliDe is a modified draft model architecture that reuses the cached keys and values from the target LLM, while CaPE is a proposal expansion method that uses the draft model's confidence scores to help select additional candidate tokens for verification.
- Self-Attention with Structural Position Representations78
This work uses dependency tree to represent the grammatical structure of a sentence, and proposes two strategies to encode the positional relationships among words in the dependency tree.
- Coverage-based Neural Machine Translation.69
Experiments show that coverage-based NMT significantly improves both alignment and translation quality over NMT without coverage.
- DukeNet68
Experimental results on two public KGC benchmarks show that Duke net significantly outperforms state-of-the-art methods in terms of both automatic and human evaluations, indicating that DukeNet enhanced by DukeL can select more appropriate knowledge and hence generate more informative and engaging responses.
- Context-aware Self-Attention Networks for Natural Language Processing67
This work casts local context modeling as a learnable Gaussian bias, which indicates the central and scope of the local region to be paid more attention and leverage the internal representations that embed sentence-level information as the global context to contextualize the transformations of query and key layers.
- Dynamic Layer Aggregation for Neural Machine Translation with Routing-by-Agreement59
This paper proposes to use routing-by-agreement strategies to aggregate layers dynamically and shows that the proposed approach consistently outperforms the strong baseline model and a representative static aggregation model.
- Translating Phrases in Neural Machine Translation58
This work proposes a method to translate phrases in NMT by integrating a phrase memory storing target phrases from a phrase-based statistical machine translation (SMT) system into the encoder-decoder architecture of NMT.
- 56
- 56
- Rejuvenating Low-Frequency Words: Making the Most of Parallel Data in Non-Autoregressive Translation55
This work directly exposes the raw data into NAT by leveraging pretraining to rejuvenate more alignments for low-frequency target words, and combines these complementary approaches as a new training strategy for further boosting NAT performance.
- Salute the Classic: Revisiting Challenges of Machine Translation in the Age of Large Language Models51
This study revisits six core challenges of Neural Machine Translation, offering insights into their ongoing relevance in the context of advanced Large Language Models, and identifies three new challenges in LLM-based translation: inference efficiency, translation of low-resource languages during pretraining, and human-aligned evaluation.
- A Comprehensive Study of GPT-4V’s Multimodal Capabilities in Medical Imaging51
This paper presents a comprehensive evaluation of GPT-4V's capabilities across diverse medical imaging tasks, including Radiology Report Generation, Medical Visual Question Answering (VQA), and Visual Grounding, and finds the limitations of conventional evaluation metrics like the BLEU score.
- Multi-Granularity Self-Attention for Neural Machine Translation51
This work presents multi-granularity self-attention (Mg-Sa): a neural network that combines multi-head self-ATTention and phrase modeling that exploits interactions among phrases to enhance the strength of structure modeling – a commonly-cited weakness of self-Attention.
- Translating Pro-Drop Languages With Reconstruction Models51
Experimental results on both Chinese-English and Japanese-English dialogue translation tasks show that the proposed approach significantly and consistently improves translation performance over a strong NMT baseline, which is directly built on the training data annotated with DPs.
- 50
- Information Aggregation for Multi-Head Attention with Routing-by-Agreement50
Experimental results on linguistic probing tasks and machine translation tasks prove the superiority of the advanced information aggregation over the standard linear transformation.
- Modeling Past and Future for Neural Machine Translation49
This work proposes a novel mechanism that separates the source information into two parts: translated Past contents and untranslated Future contents, which are modeled by two additional recurrent layers.
- Rethinking the Value of Transformer Components48
This work evaluates the impact of individual component (sub-layer) in trained Transformer models from different perspectives and proposes a new training strategy that can improves translation performance by distinguishing the unimportant components in training.
- Progressive Multi-Granularity Training for Non-Autoregressive Translation47
This study empirically shows that NAT models are prone to learn fine-grained lower-mode knowledge, such as words and phrases, compared with sentences, and proposes progressive multi-granularity training for NAT, which improves phrase translation accuracy and model reordering ability, resulting in better translation quality against strong NAT baselines.
- How Does Selective Mechanism Improve Self-Attention Networks?46
Assessing the strengths of selective SANs (SSANs), which are implemented with a flexible and universal Gumbel-Softmax, shows that SSANs consistently outperform the standard SANs.
- Towards Understanding Neural Machine Translation with Word Importance46
This work measures the word importance by attributing the NMT output to every input word through a gradient-based method, and reveals that words of certain syntactic categories have higher importance while the categories vary across language pairs, which can inspire better design principles of NMT architectures for multi-lingual translation.
- 45
- A Novel Approach to Dropped Pronoun Translation44
This work proposes a semisupervised approach to recall possibly missing pronouns in the translation of Dropped Pronouns by both extracting rules from the DP-labelled training data and translating theDP-generated input sentences.
- Context-Aware Cross-Attention for Non-Autoregressive Translation43
Experimental results show that this approach can consistently improve translation quality over strong NAT baselines and extensive analyses demonstrate that the enhanced cross-attention achieves better exploitation of source contexts by leveraging both local and global information.
- Assessing the Ability of Self-Attention Networks to Learn Word Order36
Experimental results reveal that: 1) SAN trained on word reordering detection indeed has difficulty learning the positional information even with the position embedding; and 2) SANtrained on machine translation learns better positional information than its RNN counterpart, in which position embeddedding plays a critical role.
- Dynamic Past and Future for Neural Machine Translation33
This paper proposes to model the dynamic principles by explicitly separating source words into groups of translated and untranslated contents through parts-to-wholes assignment, and achieves substantial improvements over both Rnmt and Transformer by producing more adequate translations.
- Neural Machine Translation with Adequacy-Oriented Learning31
This work proposes an adequacyoriented learning mechanism for NMT by casting translation as a stochastic policy in Reinforcement Learning (RL), where the reward is estimated by explicitly measuring translation adequacy.
- 29
- 28
- 27
- 27
- 26
- 25
- 25
- 25
- 25
- 25
- 25
- 24
- 23
- 22
- 22
- 22
- 22
- 21
- 21
- 21
- 20
- 20
- 20
- 20
- Using Syntactic Head Information in Hierarchical Phrase-Based Translation20
The Head-Driven HPB (HD-HPB) model is presented, which incorporates head information in translation rules to better capture syntax-driven information in a derivation and allows improved reordering between any two neighboring non-terminals to explore a larger reordering search space.
- DeepTheorem: Advancing LLM Reasoning for Theorem Proving Through Natural Language and Reinforcement Learning19
Extensive experimental analyses demonstrate DeepTheorem significantly improves LLM theorem-proving performance compared to existing datasets and supervised fine-tuning protocols, achieving state-of-the-art accuracy and reasoning quality.
- 19
- 19
- 18
- 18
- 18
- Dancing with Critiques: Enhancing LLM Reasoning with Stepwise Natural Language Self-Critique16
This paper proposes a novel inference-time scaling approach -- stepwise natural language self-critique (PANEL), which employs self-generated natural language critiques as feedback to guide the step-level search process, and demonstrates that PANEL significantly enhances reasoning performance, outperforming traditional scalar reward-based methods.
- 15
- 15
- 15
- 15
- 15
- Head-Driven Hierarchical Phrase-based Translation15
An extension of Chiang's hierarchical phrase-based (HPB) model is presented, called Head-Driven HPB (HD-HPB), which incorporates head information in translation rules to better capture syntax-driven information, as well as improved reordering between any two neighboring non-terminals at any stage of a derivation to explore a larger reordering search space.
- 15
- 14
- 14
- 14
- 14
- Thoughts Are All Over the Place: On the Underthinking of Long Reasoning Models13
A decoding strategy with thought switching penalty (T IP) that discourages premature transitions between thoughts, encouraging deeper exploration of each reasoning path and improves accuracy across challenging datasets without requiring model fine-tuning is proposed.
- 13
- 13
- 13
- ngram-OAXE: Phrase-Based Order-Agnostic Cross Entropy for Non-Autoregressive Machine Translation12
This work extends oaxe by only allowing reordering between ngram phrases and still requiring a strict match of word order within the phrases, and shows that ngram noaxe indeed improves the translation of ngram words, and produces more fluent translation with a better modeling of sentence structure.
- 11
- 11
- 11
- 10
- 10
- 10
- 10
- 10
- 10
- 10
- 9
- 9
- 9
- 9
- 9
- 9
- 8
- 8
- Learning to refine source representations for neural machine translation7
A novel encoder–refiner–decoder framework, which dynamically refines the source representations based on the generated target-side information at each decoding step, and a policy network to decide when to refine at specific decoding steps is proposed.
- 7
- 7
- 6
- 6
- 6
- 6
- 5
- 5
- 5
- 5
- 5
- 5
- 5
- 4
- 4
- 4
- 4
- 4
- 4
- 4
- 3
- 3
- 3
- 3
- 2
- 2
- 2
- 2
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
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 2026-10-13. 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.