arXiv subjects

Ze Chen

Publications and source records attributed to Ze Chen.

At least 55 records · Page 3Linked to original sources

arXiv2024

OPDAI at SemEval-2024 Task 6: Small LLMs can Accelerate Hallucination Detection with Weakly Supervised Data

This paper mainly describes a unified system for hallucination detection of LLMs, which wins the second prize in the model-agnostic track of the SemEval-2024 Task 6, and also achieves considerable results in the model-aware track. This task aims to detect hallucination with LLMs for three different text-generation tasks without labeled training data. We utilize prompt engineering and few-shot learning to verify the performance of different LLMs on the validation data. Then we select the LLMs with better performance to generate high-quality weakly supervised training data, which not only satisfies the consistency of different LLMs, but also satisfies the consistency of the optimal LLM with different sampling parameters. Furthermore, we finetune different LLMs by using the constructed training data, and finding that a relatively small LLM can achieve a competitive level of performance in hallucination detection, when compared to the large LLMs and the prompt-based approaches using GPT-4.

cs.CL↗

arXiv2024

Optimal Parameter and Neuron Pruning for Out-of-Distribution Detection

For a machine learning model deployed in real world scenarios, the ability of detecting out-of-distribution (OOD) samples is indispensable and challenging. Most existing OOD detection methods focused on exploring advanced training skills or training-free tricks to prevent the model from yielding overconfident confidence score for unknown samples. The training-based methods require expensive training cost and rely on OOD samples which are not always available, while most training-free methods can not efficiently utilize the prior information from the training data. In this work, we propose an \textbf{O}ptimal \textbf{P}arameter and \textbf{N}euron \textbf{P}runing (\textbf{OPNP}) approach, which aims to identify and remove those parameters and neurons that lead to over-fitting. The main method is divided into two steps. In the first step, we evaluate the sensitivity of the model parameters and neurons by averaging gradients over all training samples. In the second step, the parameters and neurons with exceptionally large or close to zero sensitivities are removed for prediction. Our proposal is training-free, compatible with other post-hoc methods, and exploring the information from all training data. Extensive experiments are performed on multiple OOD detection tasks and model architectures, showing that our proposed OPNP consistently outperforms the existing methods by a large margin.

cs.LG↗

arXiv2023

Uncertainty-aware Unsupervised Multi-Object Tracking

Without manually annotated identities, unsupervised multi-object trackers are inferior to learning reliable feature embeddings. It causes the similarity-based inter-frame association stage also be error-prone, where an uncertainty problem arises. The frame-by-frame accumulated uncertainty prevents trackers from learning the consistent feature embedding against time variation. To avoid this uncertainty problem, recent self-supervised techniques are adopted, whereas they failed to capture temporal relations. The interframe uncertainty still exists. In fact, this paper argues that though the uncertainty problem is inevitable, it is possible to leverage the uncertainty itself to improve the learned consistency in turn. Specifically, an uncertainty-based metric is developed to verify and rectify the risky associations. The resulting accurate pseudo-tracklets boost learning the feature consistency. And accurate tracklets can incorporate temporal information into spatial transformation. This paper proposes a tracklet-guided augmentation strategy to simulate tracklets' motion, which adopts a hierarchical uncertainty-based sampling mechanism for hard sample mining. The ultimate unsupervised MOT framework, namely U2MOT, is proven effective on MOT-Challenges and VisDrone-MOT benchmark. U2MOT achieves a SOTA performance among the published supervised and unsupervised trackers.

cs.CV↗

arXiv2023

Robbed withdrawal

In this article we show that Theorem 2 in Lie et al. (2023) is incorrect. Since Wombat Exchange, a decentralized exchange, is built upon Lie et al. (2023) and Theorem 2 is fundamental to Wombat Finance, we show that an undesirable phenomenon, which we call the robbed withdrawal, can happen as a consequence.

cs.CR↗

arXiv2023

Sentence-Level Relation Extraction via Contrastive Learning with Descriptive Relation Prompts

Sentence-level relation extraction aims to identify the relation between two entities for a given sentence. The existing works mostly focus on obtaining a better entity representation and adopting a multi-label classifier for relation extraction. A major limitation of these works is that they ignore background relational knowledge and the interrelation between entity types and candidate relations. In this work, we propose a new paradigm, Contrastive Learning with Descriptive Relation Prompts(CTL-DRP), to jointly consider entity information, relational knowledge and entity type restrictions. In particular, we introduce an improved entity marker and descriptive relation prompts when generating contextual embedding, and utilize contrastive learning to rank the restricted candidate relations. The CTL-DRP obtains a competitive F1-score of 76.7% on TACRED. Furthermore, the new presented paradigm achieves F1-scores of 85.8% and 91.6% on TACREV and Re-TACRED respectively, which are both the state-of-the-art performance.

cs.CL↗

arXiv2023

Enhancing Model Performance in Multilingual Information Retrieval with Comprehensive Data Engineering Techniques

In this paper, we present our solution to the Multilingual Information Retrieval Across a Continuum of Languages (MIRACL) challenge of WSDM CUP 2023\footnote{https://project-miracl.github.io/}. Our solution focuses on enhancing the ranking stage, where we fine-tune pre-trained multilingual transformer-based models with MIRACL dataset. Our model improvement is mainly achieved through diverse data engineering techniques, including the collection of additional relevant training data, data augmentation, and negative sampling. Our fine-tuned model effectively determines the semantic relevance between queries and documents, resulting in a significant improvement in the efficiency of the multilingual information retrieval process. Finally, Our team is pleased to achieve remarkable results in this challenging competition, securing 2nd place in the Surprise-Languages track with a score of 0.835 and 3rd place in the Known-Languages track with an average nDCG@10 score of 0.716 across the 16 known languages on the final leaderboard.

cs.IR↗

arXiv2023

OPD@NL4Opt: An ensemble approach for the NER task of the optimization problem

In this paper, we present an ensemble approach for the NL4Opt competition subtask 1(NER task). For this task, we first fine tune the pretrained language models based on the competition dataset. Then we adopt differential learning rates and adversarial training strategies to enhance the model generalization and robustness. Additionally, we use a model ensemble method for the final prediction, which achieves a micro-averaged F1 score of 93.3% and attains the second prize in the NER task.

cs.CL↗

arXiv2022

Multifold Cross-Validation Model Averaging for Generalized Additive Partial Linear Models

Generalized additive partial linear models (GAPLMs) are appealing for model interpretation and prediction. However, for GAPLMs, the covariates and the degree of smoothing in the nonparametric parts are often difficult to determine in practice. To address this model selection uncertainty issue, we develop a computationally feasible model averaging (MA) procedure. The model weights are data-driven and selected based on multifold cross-validation (CV) (instead of leave-one-out) for computational saving. When all the candidate models are misspecified, we show that the proposed MA estimator for GAPLMs is asymptotically optimal in the sense of achieving the lowest possible Kullback-Leibler loss. In the other scenario where the candidate model set contains at least one correct model, the weights chosen by the multifold CV are asymptotically concentrated on the correct models. As a by-product, we propose a variable importance measure to quantify the importances of the predictors in GAPLMs based on the MA weights. It is shown to be able to asymptotically identify the variables in the true model. Moreover, when the number of candidate models is very large, a model screening method is provided. Numerical experiments show the superiority of the proposed MA method over some existing model averaging and selection methods.

stat.ME↗

arXiv2022

Using Deep Mixture-of-Experts to Detect Word Meaning Shift for TempoWiC

This paper mainly describes the dma submission to the TempoWiC task, which achieves a macro-F1 score of 77.05% and attains the first place in this task. We first explore the impact of different pre-trained language models. Then we adopt data cleaning, data augmentation, and adversarial training strategies to enhance the model generalization and robustness. For further improvement, we integrate POS information and word semantic representation using a Mixture-of-Experts (MoE) approach. The experimental results show that MoE can overcome the feature overuse issue and combine the context, POS, and word semantic features well. Additionally, we use a model ensemble method for the final prediction, which has been proven effective by many research works.

cs.CL↗

arXiv2022

Variable Importance Based Interaction Modeling with an Application on Initial Spread of COVID-19 in China

Interaction selection for linear regression models with both continuous and categorical predictors is useful in many fields of modern science, yet very challenging when the number of predictors is relatively large. Existing interaction selection methods focus on finding one optimal model. While attractive properties such as consistency and oracle property have been well established for such methods, they actually may perform poorly in terms of stability for high-dimensional data, and they do not typically deal with categorical predictors. In this paper, we introduce a variable importance based interaction modeling (VIBIM) procedure for learning interactions in a linear regression model with both continuous and categorical predictors. It delivers multiple strong candidate models with high stability and interpretability. Simulation studies demonstrate its good finite sample performance. We apply the VIBIM procedure to a Corona Virus Disease 2019 (COVID-19) data used in Tian et al. (2020) and measure the effects of relevant factors, including transmission control measures on the spread of COVID-19. We show that the VIBIM approach leads to better models in terms of interpretability, stability, reliability and prediction.

stat.ME↗

arXiv2022

Corticosteroid Activation of Atlantic Sea Lamprey Corticoid Receptor: Allosteric Regulation by the N-terminal Domain

Lampreys are jawless fish that evolved about 550 million years ago at the base of the vertebrate line. Modern lampreys contain a corticoid receptor (CR), the common ancestor of the glucocorticoid receptor (GR) and mineralocorticoid receptor (MR), which first appear in cartilaginous fish, such as sharks. Until recently, 344 amino acids at the amino terminus of adult lamprey CR were not present in the lamprey CR sequence in GenBank. A search of the recently sequenced lamprey germline genome identified two CR sequences, CR1 and CR2, containing the 344 previously un-identified amino acids at the amino terminus. CR1 also contains a novel four amino acid insertion in the DNA-binding domain (DBD). We studied corticosteroid activation of CR1 and CR2 and found their strongest response was to 11-deoxycorticosterone and 11-deoxycortisol, the two circulating corticosteroids in lamprey. Based on steroid specificity, both CRs are close to elephant shark MR and distant from elephant shark GR. HEK293 cells transfected with full-length CR1 or CR2 and the MMTV promoter have about 3-fold higher steroid-mediated activation compared to HEK293 cells transfected with these CRs and the TAT3 promoter. Deletion of the amino-terminal domain (NTD) of lamprey CR1 and CR2 to form truncated CRs decreased transcriptional activation by about 70% in HEK293 cells transfected with MMTV, but increased transcription by about 6-fold in cells transfected with TAT3, indicating that the promoter has an important effect on NTD regulation of CR transcription by corticosteroids.

q-bio.BM↗

arXiv2022

A Semantic Alignment System for Multilingual Query-Product Retrieval

This paper mainly describes our winning solution (team name: www) to Amazon ESCI Challenge of KDD CUP 2022, which achieves a NDCG score of 0.9043 and wins the first place on task 1: the query-product ranking track. In this competition, participants are provided with a real-world large-scale multilingual shopping queries data set and it contains query-product pairs in English, Japanese and Spanish. Three different tasks are proposed in this competition, including ranking the results list as task 1, classifying the query/product pairs into Exact, Substitute, Complement, or Irrelevant (ESCI) categories as task 2 and identifying substitute products for a given query as task 3. We mainly focus on task 1 and propose a semantic alignment system for multilingual query-product retrieval. Pre-trained multilingual language models (LM) are adopted to get the semantic representation of queries and products. Our models are all trained with cross-entropy loss to classify the query-product pairs into ESCI 4 categories at first, and then we use weighted sum with the 4-class probabilities to get the score for ranking. To further boost the model, we also do elaborative data preprocessing, data augmentation by translation, specially handling English texts with English LMs, adversarial training with AWP and FGM, self distillation, pseudo labeling, label smoothing and ensemble. Finally, Our solution outperforms others both on public and private leaderboard.

cs.IR↗

arXiv2022

Spatial Likelihood Voting with Self-Knowledge Distillation for Weakly Supervised Object Detection

Weakly supervised object detection (WSOD), which is an effective way to train an object detection model using only image-level annotations, has attracted considerable attention from researchers. However, most of the existing methods, which are based on multiple instance learning (MIL), tend to localize instances to the discriminative parts of salient objects instead of the entire content of all objects. In this paper, we propose a WSOD framework called the Spatial Likelihood Voting with Self-knowledge Distillation Network (SLV-SD Net). In this framework, we introduce a spatial likelihood voting (SLV) module to converge region proposal localization without bounding box annotations. Specifically, in every iteration during training, all the region proposals in a given image act as voters voting for the likelihood of each category in the spatial dimensions. After dilating the alignment on the area with large likelihood values, the voting results are regularized as bounding boxes, which are then used for the final classification and localization. Based on SLV, we further propose a self-knowledge distillation (SD) module to refine the feature representations of the given image. The likelihood maps generated by the SLV module are used to supervise the feature learning of the backbone network, encouraging the network to attend to wider and more diverse areas of the image. Extensive experiments on the PASCAL VOC 2007/2012 and MS-COCO datasets demonstrate the excellent performance of SLV-SD Net. In addition, SLV-SD Net produces new state-of-the-art results on these benchmarks.

cs.CV↗

arXiv2022

Dynamic Supervisor for Cross-dataset Object Detection

The application of cross-dataset training in object detection tasks is complicated because the inconsistency in the category range across datasets transforms fully supervised learning into semi-supervised learning. To address this problem, recent studies focus on the generation of high-quality missing annotations. In this study, we first point out that it is not enough to generate high-quality annotations using a single model, which only looks once for annotations. Through detailed experimental analyses, we further conclude that hard-label training is conducive to generating high-recall annotations, while soft-label training tends to obtain high-precision annotations. Inspired by the aspects mentioned above, we propose a dynamic supervisor framework that updates the annotations multiple times through multiple-updated submodels trained using hard and soft labels. In the final generated annotations, both recall and precision improve significantly through the integration of hard-label training with soft-label training. Extensive experiments conducted on various dataset combination settings support our analyses and demonstrate the superior performance of the proposed dynamic supervisor.

cs.CV↗

arXiv2021

Attend to Who You Are: Supervising Self-Attention for Keypoint Detection and Instance-Aware Association

This paper presents a new method to solve keypoint detection and instance association by using Transformer. For bottom-up multi-person pose estimation models, they need to detect keypoints and learn associative information between keypoints. We argue that these problems can be entirely solved by Transformer. Specifically, the self-attention in Transformer measures dependencies between any pair of locations, which can provide association information for keypoints grouping. However, the naive attention patterns are still not subjectively controlled, so there is no guarantee that the keypoints will always attend to the instances to which they belong. To address it we propose a novel approach of supervising self-attention for multi-person keypoint detection and instance association. By using instance masks to supervise self-attention to be instance-aware, we can assign the detected keypoints to their corresponding instances based on the pairwise attention scores, without using pre-defined offset vector fields or embedding like CNN-based bottom-up models. An additional benefit of our method is that the instance segmentation results of any number of people can be directly obtained from the supervised attention matrix, thereby simplifying the pixel assignment pipeline. The experiments on the COCO multi-person keypoint detection challenge and person instance segmentation task demonstrate the effectiveness and simplicity of the proposed method and show a promising way to control self-attention behavior for specific purposes.

cs.CV↗

arXiv2021

Rapid Application of the Spherical Harmonic Transform via Interpolative Decomposition Butterfly Factorization

We describe an algorithm for the application of the forward and inverse spherical harmonic transforms. It is based on a new method for rapidly computing the forward and inverse associated Legendre transforms by hierarchically applying the interpolative decomposition butterfly factorization (IDBF). Experimental evidence suggests that the total running time of our method -- including all necessary precomputations -- is $\mathcal{O}(N^2 \log^3(N))$, where $N$ is the order of the transform. This is nearly asymptotically optimal. Moreover, unlike existing algorithms which are asymptotically optimal or nearly so, the constant in the running time of our algorithm is small enough to make it competitive with state-of-the-art $\mathcal{O}\left(N^3\right)$ methods at relatively small values of $N$. Numerical results are provided to demonstrate the effectiveness and numerical stability of the new framework.

math.NA↗

arXiv2021

Guiding Query Position and Performing Similar Attention for Transformer-Based Detection Heads

After DETR was proposed, this novel transformer-based detection paradigm which performs several cross-attentions between object queries and feature maps for predictions has subsequently derived a series of transformer-based detection heads. These models iterate object queries after each cross-attention. However, they don't renew the query position which indicates object queries' position information. Thus model needs extra learning to figure out the newest regions that query position should express and need more attention. To fix this issue, we propose the Guided Query Position (GQPos) method to embed the latest location information of object queries to query position iteratively. Another problem of such transformer-based detection heads is the high complexity to perform attention on multi-scale feature maps, which hinders them from improving detection performance at all scales. Therefore we propose a novel fusion scheme named Similar Attention (SiA): besides the feature maps is fused, SiA also fuse the attention weights maps to accelerate the learning of high-resolution attention weight map by well-learned low-resolution attention weight map. Our experiments show that the proposed GQPos improves the performance of a series of models, including DETR, SMCA, YoloS, and HoiTransformer and SiA consistently improve the performance of multi-scale transformer-based detection heads like DETR and HoiTransformer.

cs.CV↗

arXiv2020

SLV: Spatial Likelihood Voting for Weakly Supervised Object Detection

Based on the framework of multiple instance learning (MIL), tremendous works have promoted the advances of weakly supervised object detection (WSOD). However, most MIL-based methods tend to localize instances to their discriminative parts instead of the whole content. In this paper, we propose a spatial likelihood voting (SLV) module to converge the proposal localizing process without any bounding box annotations. Specifically, all region proposals in a given image play the role of voters every iteration during training, voting for the likelihood of each category in spatial dimensions. After dilating alignment on the area with large likelihood values, the voting results are regularized as bounding boxes, being used for the final classification and localization. Based on SLV, we further propose an end-to-end training framework for multi-task learning. The classification and localization tasks promote each other, which further improves the detection performance. Extensive experiments on the PASCAL VOC 2007 and 2012 datasets demonstrate the superior performance of SLV.

cs.CV↗