arXiv subjects

Ye Tian

Publications and source records attributed to Ye Tian.

At least 181 records · Page 10Linked to original sources

arXiv2024

Self-Consistency Boosts Calibration for Math Reasoning

Calibration, which establishes the correlation between accuracy and model confidence, is important for LLM development. We design three off-the-shelf calibration methods based on self-consistency (Wang et al., 2022) for math reasoning tasks. Evaluation on two popular benchmarks (GSM8K and MathQA) using strong open-source LLMs (Mistral and LLaMA2), our methods better bridge model confidence and accuracy than existing methods based on p(True) (Kadavath et al., 2022) or logit (Kadavath et al., 2022).

cs.CL↗

arXiv2024

VQGraph: Rethinking Graph Representation Space for Bridging GNNs and MLPs

GNN-to-MLP distillation aims to utilize knowledge distillation (KD) to learn computationally-efficient multi-layer perceptron (student MLP) on graph data by mimicking the output representations of teacher GNN. Existing methods mainly make the MLP to mimic the GNN predictions over a few class labels. However, the class space may not be expressive enough for covering numerous diverse local graph structures, thus limiting the performance of knowledge transfer from GNN to MLP. To address this issue, we propose to learn a new powerful graph representation space by directly labeling nodes' diverse local structures for GNN-to-MLP distillation. Specifically, we propose a variant of VQ-VAE to learn a structure-aware tokenizer on graph data that can encode each node's local substructure as a discrete code. The discrete codes constitute a codebook as a new graph representation space that is able to identify different local graph structures of nodes with the corresponding code indices. Then, based on the learned codebook, we propose a new distillation target, namely soft code assignments, to directly transfer the structural knowledge of each node from GNN to MLP. The resulting framework VQGraph achieves new state-of-the-art performance on GNN-to-MLP distillation in both transductive and inductive settings across seven graph datasets. We show that VQGraph with better performance infers faster than GNNs by 828x, and also achieves accuracy improvement over GNNs and stand-alone MLPs by 3.90% and 28.05% on average, respectively. Code: https://github.com/YangLing0818/VQGraph.

cs.LG↗

arXiv2024

Collaborative decoding of critical tokens for boosting factuality of large language models

The most common training pipeline for large language models includes pretraining, finetuning and aligning phases, with their respective resulting models, such as the pretrained model and the finetuned model. Finetuned and aligned models show improved abilities of instruction following and safe generation, however their abilities to stay factual about the world are impacted by the finetuning process. Furthermore, the common practice of using sampling during generation also increases chances of hallucination. In this work, we introduce a collaborative decoding framework to harness the high factuality within pretrained models through the concept of critical tokens. We first design a critical token classifier to decide which model to use for the next token, and subsequently generates the next token using different decoding strategies. Experiments with different models and datasets show that our decoding framework is able to reduce model hallucination significantly, showcasing the importance of the collaborative decoding framework.

cs.CL↗

arXiv2024

Fine-Grained Self-Endorsement Improves Factuality and Reasoning

This work studies improving large language model (LLM) generations at inference time by mitigating fact-conflicting hallucinations. Particularly, we propose a self-endorsement framework that leverages the fine-grained fact-level comparisons across multiple sampled responses. Compared with prior ensemble methods (Wang et al., 2022;Chen et al., 2023)) that perform response-level selection, our approach can better alleviate hallucinations, especially for longform generation tasks. Our approach can broadly benefit smaller and open-source LLMs as it mainly conducts simple content-based comparisons. Experiments on Biographies show that our method can effectively improve the factuality of generations with simple and intuitive prompts across different scales of LLMs. Besides, comprehensive analyses on TriviaQA and GSM8K demonstrate the potential of self-endorsement for broader application.

cs.CL↗

arXiv2023

FCDNet: Frequency-Guided Complementary Dependency Modeling for Multivariate Time-Series Forecasting

Multivariate time-series (MTS) forecasting is a challenging task in many real-world non-stationary dynamic scenarios. In addition to intra-series temporal signals, the inter-series dependency also plays a crucial role in shaping future trends. How to enable the model's awareness of dependency information has raised substantial research attention. Previous approaches have either presupposed dependency constraints based on domain knowledge or imposed them using real-time feature similarity. However, MTS data often exhibit both enduring long-term static relationships and transient short-term interactions, which mutually influence their evolving states. It is necessary to recognize and incorporate the complementary dependencies for more accurate MTS prediction. The frequency information in time series reflects the evolutionary rules behind complex temporal dynamics, and different frequency components can be used to well construct long-term and short-term interactive dependency structures between variables. To this end, we propose FCDNet, a concise yet effective framework for multivariate time-series forecasting. Specifically, FCDNet overcomes the above limitations by applying two light-weight dependency constructors to help extract long- and short-term dependency information adaptively from multi-level frequency patterns. With the growth of input variables, the number of trainable parameters in FCDNet only increases linearly, which is conducive to the model's scalability and avoids over-fitting. Additionally, adopting a frequency-based perspective can effectively mitigate the influence of noise within MTS data, which helps capture more genuine dependencies. The experimental results on six real-world datasets from multiple fields show that FCDNet significantly exceeds strong baselines, with an average improvement of 6.82% on MAE, 4.98% on RMSE, and 4.91% on MAPE.

cs.LG↗

arXiv2023

On semi-supervised estimation using exponential tilt mixture models

Consider a semi-supervised setting with a labeled dataset of binary responses and predictors and an unlabeled dataset with only the predictors. Logistic regression is equivalent to an exponential tilt model in the labeled population. For semi-supervised estimation, we develop further analysis and understanding of a statistical approach using exponential tilt mixture (ETM) models and maximum nonparametric likelihood estimation, while allowing that the class proportions may differ between the unlabeled and labeled data. We derive asymptotic properties of ETM-based estimation and demonstrate improved efficiency over supervised logistic regression in a random sampling setup and an outcome-stratified sampling setup previously used. Moreover, we reconcile such efficiency improvement with the existing semiparametric efficiency theory when the class proportions in the unlabeled and labeled data are restricted to be the same. We also provide a simulation study to numerically illustrate our theoretical findings.

stat.ML↗

arXiv2023

On the realisation problem for mapping degree sets

The set of degrees of maps $D(M,N)$, where $M,N$ are closed oriented $n$-manifolds, always contains $0$ and the set of degrees of self-maps $D(M)$ always contains $0$ and $1$. Also, if $a,b\in D(M)$, then $ab\in D(M)$; a set $A\subseteq\mathbb Z$ so that $ab\in A$ for each $a,b\in A$ is called multiplicative. On the one hand, not every infinite set of integers (containing $0$) is a mapping degree set [NWW] and, on the other hand, every finite set of integers (containing $0$) is the mapping degree set of some $3$-manifolds [CMV]. We show the following: (i) Not every multiplicative set $A$ containing $0,1$ is a self-mapping degree set. (ii) For each $n\in\mathbb N$ and $k\geq3$, every $D(M,N)$ for $n$-manifolds $M$ and $N$ is $D(P,Q)$ for some $(n+k)$-manifolds $P$ and $Q$. As a consequence of (ii) and [CMV], every finite set of integers (containing $0$) is the mapping degree set of some $n$-manifolds for all $n\neq 1,2,4,5$.

math.GT↗

arXiv2023

Learning From Free-Text Human Feedback -- Collect New Datasets Or Extend Existing Ones?

Learning from free-text human feedback is essential for dialog systems, but annotated data is scarce and usually covers only a small fraction of error types known in conversational AI. Instead of collecting and annotating new datasets from scratch, recent advances in synthetic dialog generation could be used to augment existing dialog datasets with the necessary annotations. However, to assess the feasibility of such an effort, it is important to know the types and frequency of free-text human feedback included in these datasets. In this work, we investigate this question for a variety of commonly used dialog datasets, including MultiWoZ, SGD, BABI, PersonaChat, Wizards-of-Wikipedia, and the human-bot split of the Self-Feeding Chatbot. Using our observations, we derive new taxonomies for the annotation of free-text human feedback in dialogs and investigate the impact of including such data in response generation for three SOTA language generation models, including GPT-2, LLAMA, and Flan-T5. Our findings provide new insights into the composition of the datasets examined, including error types, user response types, and the relations between them.

cs.CL↗

arXiv2023

Incorporating Domain Knowledge Graph into Multimodal Movie Genre Classification with Self-Supervised Attention and Contrastive Learning

Multimodal movie genre classification has always been regarded as a demanding multi-label classification task due to the diversity of multimodal data such as posters, plot summaries, trailers and metadata. Although existing works have made great progress in modeling and combining each modality, they still face three issues: 1) unutilized group relations in metadata, 2) unreliable attention allocation, and 3) indiscriminative fused features. Given that the knowledge graph has been proven to contain rich information, we present a novel framework that exploits the knowledge graph from various perspectives to address the above problems. As a preparation, the metadata is processed into a domain knowledge graph. A translate model for knowledge graph embedding is adopted to capture the relations between entities. Firstly we retrieve the relevant embedding from the knowledge graph by utilizing group relations in metadata and then integrate it with other modalities. Next, we introduce an Attention Teacher module for reliable attention allocation based on self-supervised learning. It learns the distribution of the knowledge graph and produces rational attention weights. Finally, a Genre-Centroid Anchored Contrastive Learning module is proposed to strengthen the discriminative ability of fused features. The embedding space of anchors is initialized from the genre entities in the knowledge graph. To verify the effectiveness of our framework, we collect a larger and more challenging dataset named MM-IMDb 2.0 compared with the MM-IMDb dataset. The experimental results on two datasets demonstrate that our model is superior to the state-of-the-art methods. We will release the code in the near future.

cs.AI↗

arXiv2023

Two-level approximation of transmons in quantum quench experiments

Quantum quench is a typical protocol in the study of nonequilibrium dynamics of quantum many-body systems. Recently, a number of experiments with superconducting transmon qubits are reported, in which the spin and hard-core boson models with two energy levels on individual sites are used. The transmons are a multilevel system and the coupled qubits are governed by the Bose-Hubbard model. How well they can be approximated by a two-level system has been discussed and analysed in different ways for specific experiments in the literature. Here, we numerically investigate the accuracy and validity of the two-level approximation for the multilevel transmons based on the concept of Loschmidt echo. Using this method, we are able to calculate the fidelity decay (i.e., the time-dependent overlap of evolving wave functions) due to the state leakage to transmon high energy levels. We present the results for different system Hamiltonians with various initial states, qubit coupling strength, and external driving, and for two kinds of quantum quench experiments with time reversal and time evolution in one direction. We show quantitatively the extent to which the fidelity decays with time for changing coupling strength (or on-site interaction over coupling strength) and filled particle number or locations in the initial states under specific system Hamiltonians, which may serve as a way for assessing the two-level approximation of transmons. Finally, we compare our results with the reported experiments using transmon qubits.

quant-ph↗

arXiv2023

Evolutionary Neural Architecture Search for Transformer in Knowledge Tracing

Knowledge tracing (KT) aims to trace students' knowledge states by predicting whether students answer correctly on exercises. Despite the excellent performance of existing Transformer-based KT approaches, they are criticized for the manually selected input features for fusion and the defect of single global context modelling to directly capture students' forgetting behavior in KT, when the related records are distant from the current record in terms of time. To address the issues, this paper first considers adding convolution operations to the Transformer to enhance its local context modelling ability used for students' forgetting behavior, then proposes an evolutionary neural architecture search approach to automate the input feature selection and automatically determine where to apply which operation for achieving the balancing of the local/global context modelling. In the search space, the original global path containing the attention module in Transformer is replaced with the sum of a global path and a local path that could contain different convolutions, and the selection of input features is also considered. To search the best architecture, we employ an effective evolutionary algorithm to explore the search space and also suggest a search space reduction strategy to accelerate the convergence of the algorithm. Experimental results on the two largest and most challenging education datasets demonstrate the effectiveness of the architecture found by the proposed approach.

cs.NE↗

arXiv2023

Observation of universal dissipative dynamics in strongly correlated quantum gas

Dissipation is unavoidable in quantum systems. It usually induces decoherences and changes quantum correlations. To access the information of strongly correlated quantum matters, one has to overcome or suppress dissipation to extract out the underlying quantum phenomena. However, here we find an opposite effect that dissipation can be utilized as a powerful tool to probe the intrinsic correlations of quantum many-body systems. Applying highly-controllable dissipation in ultracold atomic systems, we observe a universal dissipative dynamics in strongly correlated one-dimensional quantum gases. The total particle number of this system follows a universal stretched-exponential decay, and the stretched exponent measures the anomalous dimension of the spectral function, a critical exponent characterizing strong quantum fluctuations of this system. This method could have broad applications in detecting strongly correlated features, including spin-charge separations and Fermi arcs in quantum materials.

cond-mat.quant-gas↗

arXiv2023

Stabilizing RLHF through Advantage Model and Selective Rehearsal

Large Language Models (LLMs) have revolutionized natural language processing, yet aligning these models with human values and preferences using RLHF remains a significant challenge. This challenge is characterized by various instabilities, such as reward hacking and catastrophic forgetting. In this technical report, we propose two innovations to stabilize RLHF training: 1) Advantage Model, which directly models advantage score i.e., extra reward compared to the expected rewards and regulates score distributions across tasks to prevent reward hacking. 2) Selective Rehearsal, which mitigates catastrophic forgetting by strategically selecting data for PPO training and knowledge rehearsing. Our experimental analysis on public and proprietary datasets reveals that the proposed methods not only increase stability in RLHF training but also achieve higher reward scores and win rates.

cs.CL↗

arXiv2023

Simulating Chern insulators on a superconducting quantum processor

The quantum Hall effect, fundamental in modern condensed matter physics, continuously inspires new theories and predicts emergent phases of matter. Here we experimentally demonstrate three types of Chern insulators with synthetic dimensions on a programable 30-qubit-ladder superconducting processor. We directly measure the band structures of the 2D Chern insulator along synthetic dimensions with various configurations of Aubry-André-Harper chains and observe dynamical localisation of edge excitations. With these two signatures of topology, our experiments implement the bulk-edge correspondence in the synthetic 2D Chern insulator. Moreover, we simulate two different bilayer Chern insulators on the ladder-type superconducting processor. With the same and opposite periodically modulated on-site potentials for two coupled chains, we simulate topologically nontrivial edge states with zero Hall conductivity and a Chern insulator with higher Chern numbers, respectively. Our work shows the potential of using superconducting qubits for investigating different intriguing topological phases of quantum matter.

quant-ph↗

arXiv2023

RestGPT: Connecting Large Language Models with Real-World RESTful APIs

Tool-augmented large language models (LLMs) have achieved remarkable progress in tackling a broad range of tasks. However, existing methods are mainly restricted to specifically designed tools and fail to fulfill complex instructions, having great limitations when confronted with real-world scenarios. In this paper, we explore a more realistic scenario by connecting LLMs with RESTful APIs, which adhere to the widely adopted REST software architectural style for web service development. To address the practical challenges of tackling complex instructions, we propose RestGPT, which exploits the power of LLMs and conducts a coarse-to-fine online planning mechanism to enhance the abilities of task decomposition and API selection. RestGPT also contains an API executor tailored for calling RESTful APIs, which can meticulously formulate parameters and parse API responses. To fully evaluate the performance of RestGPT, we propose RestBench, a high-quality benchmark which consists of two real-world scenarios and human-annotated instructions with gold solution paths. Experiments show that RestGPT is able to achieve impressive results in complex tasks and has strong robustness, which paves a new way towards AGI. RestGPT and RestBench is publicly available at https://restgpt.github.io/.

cs.CL↗

arXiv2023

EQ-Net: Elastic Quantization Neural Networks

Current model quantization methods have shown their promising capability in reducing storage space and computation complexity. However, due to the diversity of quantization forms supported by different hardware, one limitation of existing solutions is that usually require repeated optimization for different scenarios. How to construct a model with flexible quantization forms has been less studied. In this paper, we explore a one-shot network quantization regime, named Elastic Quantization Neural Networks (EQ-Net), which aims to train a robust weight-sharing quantization supernet. First of all, we propose an elastic quantization space (including elastic bit-width, granularity, and symmetry) to adapt to various mainstream quantitative forms. Secondly, we propose the Weight Distribution Regularization Loss (WDR-Loss) and Group Progressive Guidance Loss (GPG-Loss) to bridge the inconsistency of the distribution for weights and output logits in the elastic quantization space gap. Lastly, we incorporate genetic algorithms and the proposed Conditional Quantization-Aware Accuracy Predictor (CQAP) as an estimator to quickly search mixed-precision quantized neural networks in supernet. Extensive experiments demonstrate that our EQ-Net is close to or even better than its static counterparts as well as state-of-the-art robust bit-width methods. Code can be available at \href{https://github.com/xuke225/EQ-Net.git}{https://github.com/xuke225/EQ-Net}.

cs.CV↗

arXiv2023

Improving Social Media Popularity Prediction with Multiple Post Dependencies

Social Media Popularity Prediction has drawn a lot of attention because of its profound impact on many different applications, such as recommendation systems and multimedia advertising. Despite recent efforts to leverage the content of social media posts to improve prediction accuracy, many existing models fail to fully exploit the multiple dependencies between posts, which are important to comprehensively extract content information from posts. To tackle this problem, we propose a novel prediction framework named Dependency-aware Sequence Network (DSN) that exploits both intra- and inter-post dependencies. For intra-post dependency, DSN adopts a multimodal feature extractor with an efficient fine-tuning strategy to obtain task-specific representations from images and textual information of posts. For inter-post dependency, DSN uses a hierarchical information propagation method to learn category representations that could better describe the difference between posts. DSN also exploits recurrent networks with a series of gating layers for more flexible local temporal processing abilities and multi-head attention for long-term dependencies. The experimental results on the Social Media Popularity Dataset demonstrate the superiority of our method compared to existing state-of-the-art models.

cs.MM↗

arXiv2023

Quantum simulation of topological zero modes on a 41-qubit superconducting processor

Quantum simulation of different exotic topological phases of quantum matter on a noisy intermediate-scale quantum (NISQ) processor is attracting growing interest. Here, we develop a one-dimensional 43-qubit superconducting quantum processor, named as Chuang-tzu, to simulate and characterize emergent topological states. By engineering diagonal Aubry-Andr$\acute{\mathrm{e}}$-Harper (AAH) models, we experimentally demonstrate the Hofstadter butterfly energy spectrum. Using Floquet engineering, we verify the existence of the topological zero modes in the commensurate off-diagonal AAH models, which have never been experimentally realized before. Remarkably, the qubit number over 40 in our quantum processor is large enough to capture the substantial topological features of a quantum system from its complex band structure, including Dirac points, the energy gap's closing, the difference between even and odd number of sites, and the distinction between edge and bulk states. Our results establish a versatile hybrid quantum simulation approach to exploring quantum topological systems in the NISQ era.

quant-ph↗