arXiv subjects

Hao Peng

Publications and source records attributed to Hao Peng.

At least 235 records · Page 13Linked to original sources

arXiv2024

Event-level Knowledge Editing

Knowledge editing aims at updating knowledge of large language models (LLMs) to prevent them from becoming outdated. Existing work edits LLMs at the level of factual knowledge triplets. However, natural knowledge updates in the real world come from the occurrences of new events rather than direct changes in factual triplets. In this paper, we propose a new task setting: event-level knowledge editing, which directly edits new events into LLMs and improves over conventional triplet-level editing on (1) Efficiency. A single event edit leads to updates in multiple entailed knowledge triplets. (2) Completeness. Beyond updating factual knowledge, event-level editing also requires considering the event influences and updating LLMs' knowledge about future trends. We construct a high-quality event-level editing benchmark ELKEN, consisting of 1,515 event edits, 6,449 questions about factual knowledge, and 10,150 questions about future tendencies. We systematically evaluate the performance of various knowledge editing methods and LLMs on this benchmark. We find that ELKEN poses significant challenges to existing knowledge editing approaches. Our codes and dataset are publicly released to facilitate further research.

cs.CL↗

arXiv2024

Unsupervised Social Bot Detection via Structural Information Theory

Research on social bot detection plays a crucial role in maintaining the order and reliability of information dissemination while increasing trust in social interactions. The current mainstream social bot detection models rely on black-box neural network technology, e.g., Graph Neural Network, Transformer, etc., which lacks interpretability. In this work, we present UnDBot, a novel unsupervised, interpretable, yet effective and practical framework for detecting social bots. This framework is built upon structural information theory. We begin by designing three social relationship metrics that capture various aspects of social bot behaviors: Posting Type Distribution, Posting Influence, and Follow-to-follower Ratio. Three new relationships are utilized to construct a new, unified, and weighted social multi-relational graph, aiming to model the relevance of social user behaviors and discover long-distance correlations between users. Second, we introduce a novel method for optimizing heterogeneous structural entropy. This method involves the personalized aggregation of edge information from the social multi-relational graph to generate a two-dimensional encoding tree. The heterogeneous structural entropy facilitates decoding of the substantial structure of the social bots network and enables hierarchical clustering of social bots. Thirdly, a new community labeling method is presented to distinguish social bot communities by computing the user's stationary distribution, measuring user contributions to network structure, and counting the intensity of user aggregation within the community. Compared with ten representative social bot detection approaches, comprehensive experiments demonstrate the advantages of effectiveness and interpretability of UnDBot on four real social network datasets.

cs.SI↗

arXiv2024

DEFM: Delay E mbedding based Forecast Machine for Time Series Forecasting by Spatiotemporal Information Transformation

Making accurate forecasts for a complex system is a challenge in various practical applications. The major difficulty in solving such a problem concerns nonlinear spatiotemporal dynamics with time-varying characteristics. Takens' delay embedding theory provides a way to transform high-dimensional spatial information into temporal information. In this work, by combining delay embedding theory and deep learning techniques, we propose a novel framework, Delay-Embedding-based Forecast Machine (DEFM), to predict the future values of a target variable in a self-supervised and multistep-ahead manner based on high-dimensional observations. With a three-module spatiotemporal architecture, the DEFM leverages deep neural networks to effectively extract both the spatially and temporally associated information from the observed time series even with time-varying parameters or additive noise. The DEFM can accurately predict future information by transforming spatiotemporal information to the delay embeddings of a target variable. The efficacy and precision of the DEFM are substantiated through applications in three spatiotemporally chaotic systems: a 90-dimensional (90D) coupled Lorenz system, the Lorenz 96 system, and the Kuramoto-Sivashinsky (KS) equation with inhomogeneity. Additionally, the performance of the DEFM is evaluated on six real-world datasets spanning various fields. Comparative experiments with five prediction methods illustrate the superiority and robustness of the DEFM and show the great potential of the DEFM in temporal information mining and forecasting

eess.SP↗

arXiv2024

Advancing LLM Reasoning Generalists with Preference Trees

We introduce Eurus, a suite of large language models (LLMs) optimized for reasoning. Finetuned from Mistral-7B and CodeLlama-70B, Eurus models achieve state-of-the-art results among open-source models on a diverse set of benchmarks covering mathematics, code generation, and logical reasoning problems. Notably, Eurus-70B beats GPT-3.5 Turbo in reasoning through a comprehensive benchmarking across 12 tests covering five tasks, and achieves a 33.3% pass@1 accuracy on LeetCode and 32.6% on TheoremQA, two challenging benchmarks, substantially outperforming existing open-source models by margins more than 13.3%. The strong performance of Eurus can be primarily attributed to UltraInteract, our newly-curated large-scale, high-quality alignment dataset specifically designed for complex reasoning tasks. UltraInteract can be used in both supervised fine-tuning and preference learning. For each instruction, it includes a preference tree consisting of (1) reasoning chains with diverse planning strategies in a unified format, (2) multi-turn interaction trajectories with the environment and the critique, and (3) pairwise data to facilitate preference learning. UltraInteract allows us to conduct an in-depth exploration of preference learning for reasoning tasks. Our investigation reveals that some well-established preference learning algorithms may be less suitable for reasoning tasks compared to their effectiveness in general conversations. Inspired by this, we derive a novel reward modeling objective which, together with UltraInteract, leads to a strong reward model.

cs.AI↗

arXiv2024

Instruction-based Hypergraph Pretraining

Pretraining has been widely explored to augment the adaptability of graph learning models to transfer knowledge from large datasets to a downstream task, such as link prediction or classification. However, the gap between training objectives and the discrepancy between data distributions in pretraining and downstream tasks hinders the transfer of the pretrained knowledge. Inspired by instruction-based prompts widely used in pretrained language models, we introduce instructions into graph pretraining. In this paper, we propose a novel pretraining framework named Instruction-based Hypergraph Pretraining. To overcome the discrepancy between pretraining and downstream tasks, text-based instructions are applied to provide explicit guidance on specific tasks for representation learning. Compared to learnable prompts, whose effectiveness depends on the quality and the diversity of training data, text-based instructions intrinsically encapsulate task information and support the model to generalize beyond the structure seen during pretraining. To capture high-order relations with task information in a context-aware manner, a novel prompting hypergraph convolution layer is devised to integrate instructions into information propagation in hypergraphs. Extensive experiments conducted on three public datasets verify the superiority of IHP in various scenarios.

cs.IR↗

arXiv2024

Pre-training Language Model Incorporating Domain-specific Heterogeneous Knowledge into A Unified Representation

Existing technologies expand BERT from different perspectives, e.g. designing different pre-training tasks, different semantic granularities, and different model architectures. Few models consider expanding BERT from different text formats. In this paper, we propose a heterogeneous knowledge language model (\textbf{HKLM}), a unified pre-trained language model (PLM) for all forms of text, including unstructured text, semi-structured text, and well-structured text. To capture the corresponding relations among these multi-format knowledge, our approach uses masked language model objective to learn word knowledge, uses triple classification objective and title matching objective to learn entity knowledge and topic knowledge respectively. To obtain the aforementioned multi-format text, we construct a corpus in the tourism domain and conduct experiments on 5 tourism NLP datasets. The results show that our approach outperforms the pre-training of plain text using only 1/4 of the data. We further pre-train the domain-agnostic HKLM and achieve performance gains on the XNLI dataset.

cs.CL↗

arXiv2024

Language Models Hallucinate, but May Excel at Fact Verification

Recent progress in natural language processing (NLP) owes much to remarkable advances in large language models (LLMs). Nevertheless, LLMs frequently "hallucinate," resulting in non-factual outputs. Our carefully-designed human evaluation substantiates the serious hallucination issue, revealing that even GPT-3.5 produces factual outputs less than 25% of the time. This underscores the importance of fact verifiers in order to measure and incentivize progress. Our systematic investigation affirms that LLMs can be repurposed as effective fact verifiers with strong correlations with human judgments. Surprisingly, FLAN-T5-11B, the least factual generator in our study, performs the best as a fact verifier, even outperforming more capable LLMs like GPT3.5 and ChatGPT. Delving deeper, we analyze the reliance of these LLMs on high-quality evidence, as well as their deficiencies in robustness and generalization ability. Our study presents insights for developing trustworthy generation models.

cs.CL↗

arXiv2024

LeTI: Learning to Generate from Textual Interactions

Fine-tuning pre-trained language models (LMs) is essential for enhancing their capabilities. Existing techniques commonly fine-tune on input-output pairs (e.g., instruction tuning) or with numerical rewards that gauge the output quality (e.g., RLHF). We explore LMs' potential to learn from textual interactions (LETI) that not only check their correctness with binary labels but also pinpoint and explain errors in their outputs through textual feedback. Our focus is the code generation task, where the model produces code based on natural language instructions. This setting invites a natural and scalable way to acquire textual feedback: the error messages and stack traces from code execution using a Python interpreter. LETI iteratively fine-tunes the model, using the LM objective, on a concatenation of natural language instructions, LM-generated programs, and textual feedback. Prepended to this fine-tuning text, a binary reward token is used to differentiate correct and buggy solutions. LETI requires no ground-truth outputs for training and even outperforms a fine-tuned baseline that does. LETI not only improves the performance of LMs on a code generation dataset MBPP, but also generalizes to other datasets. Trained on MBPP, it achieves comparable or better performance than the base LMs on unseen problems in HumanEval. Furthermore, compared to binary feedback, we observe that textual feedback leads to improved generation quality and sample efficiency, achieving the same performance with fewer than half of the gradient steps. LETI is equally applicable in natural language tasks when they can be formulated as code generation, which we empirically verified on event argument extraction.

cs.CL↗

arXiv2024

CRAFT: Customizing LLMs by Creating and Retrieving from Specialized Toolsets

Large language models (LLMs) are often augmented with tools to solve complex tasks. By generating code snippets and executing them through task-specific Application Programming Interfaces (APIs), they can offload certain functions to dedicated external modules, such as image encoding and performing calculations. However, most existing approaches to augment LLMs with tools are constrained by general-purpose APIs and lack the flexibility for tailoring them to specific tasks. In this work, we present CRAFT, a general tool creation and retrieval framework for LLMs. It creates toolsets specifically curated for the tasks and equips LLMs with a component that retrieves tools from these sets to enhance their capability to solve complex tasks. For each task, we collect specific code solutions by prompting GPT-4 to solve the training examples. Following a validation step ensuring the correctness, these solutions are abstracted into code snippets to enhance reusability, and deduplicated for higher quality. At inference time, the language model retrieves snippets from the toolsets and then executes them or generates the output conditioning on the retrieved snippets. Our method is designed to be flexible and offers a plug-and-play approach to adapt off-the-shelf LLMs to unseen domains and modalities, without any finetuning. Experiments on vision-language, tabular processing, and mathematical reasoning tasks show that our approach achieves substantial improvements compared to strong baselines. In addition, our in-depth analysis reveals that: (1) consistent performance improvement can be achieved by scaling up the number of tools and the capability of the backbone models; (2) each component of our approach contributes to the performance gains; (3) the created tools are well-structured and reliable with low complexity and atomicity. The code is available at https://github.com/lifan-yuan/CRAFT.

cs.CL↗

arXiv2024

MINT: Evaluating LLMs in Multi-turn Interaction with Tools and Language Feedback

To solve complex tasks, large language models (LLMs) often require multiple rounds of interactions with the user, sometimes assisted by external tools. However, current evaluation protocols often emphasize benchmark performance with single-turn exchanges, neglecting the nuanced interactions among the user, LLMs, and external tools, while also underestimating the importance of natural language feedback from users. These oversights contribute to discrepancies between research benchmark evaluations and real-world use cases. We introduce MINT, a benchmark that evaluates LLMs' ability to solve tasks with multi-turn interactions by (1) using tools and (2) leveraging natural language feedback. To ensure reproducibility, we provide an evaluation framework where LLMs can access tools by executing Python code and receive users' natural language feedback simulated by GPT-4. We repurpose a diverse set of established evaluation datasets focusing on reasoning, coding, and decision-making and carefully curate them into a compact subset for efficient evaluation. Our analysis of 20 open- and closed-source LLMs offers intriguing findings. (a) LLMs generally benefit from tools and language feedback, with performance gains (absolute, same below) of 1-8% for each turn of tool use and 2-17% with natural language feedback. (b) Better single-turn performance does not guarantee better multi-turn performance. (c) Surprisingly, on the LLMs evaluated, supervised instruction-finetuning (SIFT) and reinforcement learning from human feedback (RLHF) generally hurt multi-turn capabilities. We expect MINT can help measure progress and incentivize research in improving LLMs' capabilities in multi-turn interactions, especially for open-source communities where multi-turn human evaluation can be less accessible compared to commercial LLMs with a larger user base.

cs.CL↗

arXiv2024

TRAM: Bridging Trust Regions and Sharpness Aware Minimization

Sharpness-aware minimization (SAM) reports improving domain generalization by reducing the loss surface curvature in the parameter space. However, generalization during fine-tuning is often more dependent on the transferability of representations in the function space. Trust-region methods (TR) target this goal by regularizing representation curvature to reduce catastrophic forgetting of pre-trained task-agnostic information while adopting task-specific skills. We consider unifying these strategies for low curvature in both parameter space and function space to improve out-of-domain (OOD) generalization. We propose Trust Region Aware Minimization (TRAM), a SAM algorithm fine-tuning for low parameter sharpness and smooth, informative representations preserving pre-trained structure. TRAM uses a trust region bound to inform the SAM adversarial neighborhood, introducing an awareness of function curvature within optimization for flatter minima. We empirically validate TRAM in vision (cross-dataset adaptation) and text (OOD language modeling, zero-shot cross-lingual transfer) tasks where robust domain transfer and representation generality are critical. TRAM outperforms SAM- and TR-based optimization across all tasks, notably surpassing competing methods for hard transfer between anticorrelated domains. TRAM establishes a novel standard in fine-tuning for domain-generalizable models with minimal additional computation over previous sharpness-aware methods.

cs.LG↗

arXiv2024

Understanding the PULSAR Effect in Combined Radiotherapy and Immunotherapy through Attention Mechanisms with a Transformer Model

PULSAR (personalized, ultra-fractionated stereotactic adaptive radiotherapy) is the adaptation of stereotactic ablative radiotherapy towards personalized cancer management. For the first time, we applied a transformer-based attention mechanism to investigate the underlying interactions between combined PULSAR and PD-L1 blockade immunotherapy based on a murine cancer model (Lewis Lung Carcinoma, LLC). The proposed approach is able to predict the trend of tumor volume change semi-quantitatively, and excels in identifying the potential causal relationships through both self-attention and cross-attention scores.

physics.med-ph↗

arXiv2024

Poincaré Differential Privacy for Hierarchy-Aware Graph Embedding

Hierarchy is an important and commonly observed topological property in real-world graphs that indicate the relationships between supervisors and subordinates or the organizational behavior of human groups. As hierarchy is introduced as a new inductive bias into the Graph Neural Networks (GNNs) in various tasks, it implies latent topological relations for attackers to improve their inference attack performance, leading to serious privacy leakage issues. In addition, existing privacy-preserving frameworks suffer from reduced protection ability in hierarchical propagation due to the deficiency of adaptive upper-bound estimation of the hierarchical perturbation boundary. It is of great urgency to effectively leverage the hierarchical property of data while satisfying privacy guarantees. To solve the problem, we propose the Poincaré Differential Privacy framework, named PoinDP, to protect the hierarchy-aware graph embedding based on hyperbolic geometry. Specifically, PoinDP first learns the hierarchy weights for each entity based on the Poincaré model in hyperbolic space. Then, the Personalized Hierarchy-aware Sensitivity is designed to measure the sensitivity of the hierarchical structure and adaptively allocate the privacy protection strength. Besides, the Hyperbolic Gaussian Mechanism (HGM) is proposed to extend the Gaussian mechanism in Euclidean space to hyperbolic space to realize random perturbations that satisfy differential privacy under the hyperbolic space metric. Extensive experiment results on five real-world datasets demonstrate the proposed PoinDP's advantages of effective privacy protection while maintaining good performance on the node classification task.

cs.LG↗

arXiv2024

Examining LLMs' Uncertainty Expression Towards Questions Outside Parametric Knowledge

Can large language models (LLMs) express their uncertainty in situations where they lack sufficient parametric knowledge to generate reasonable responses? This work aims to systematically investigate LLMs' behaviors in such situations, emphasizing the trade-off between honesty and helpfulness. To tackle the challenge of precisely determining LLMs' knowledge gaps, we diagnostically create unanswerable questions containing non-existent concepts or false premises, ensuring that they are outside the LLMs' vast training data. By compiling a benchmark, UnknownBench, which consists of both unanswerable and answerable questions, we quantitatively evaluate the LLMs' performance in maintaining honesty while being helpful. Using a model-agnostic unified confidence elicitation approach, we observe that most LLMs fail to consistently refuse or express uncertainty towards questions outside their parametric knowledge, although instruction fine-tuning and alignment techniques can provide marginal enhancements. Moreover, LLMs' uncertainty expression does not always stay consistent with the perceived confidence of their textual outputs.

cs.CL↗

arXiv2024

Data Engineering for Scaling Language Models to 128K Context

We study the continual pretraining recipe for scaling language models' context lengths to 128K, with a focus on data engineering. We hypothesize that long context modeling, in particular \textit{the ability to utilize information at arbitrary input locations}, is a capability that is mostly already acquired through large-scale pretraining, and that this capability can be readily extended to contexts substantially longer than seen during training~(e.g., 4K to 128K) through lightweight continual pretraining on appropriate data mixture. We investigate the \textit{quantity} and \textit{quality} of the data for continual pretraining: (1) for quantity, we show that 500 million to 5 billion tokens are enough to enable the model to retrieve information anywhere within the 128K context; (2) for quality, our results equally emphasize \textit{domain balance} and \textit{length upsampling}. Concretely, we find that naively upsampling longer data on certain domains like books, a common practice of existing work, gives suboptimal performance, and that a balanced domain mixture is important. We demonstrate that continual pretraining of the full model on 1B-5B tokens of such data is an effective and affordable strategy for scaling the context length of language models to 128K. Our recipe outperforms strong open-source long-context models and closes the gap to frontier models like GPT-4 128K.

cs.CL↗

arXiv2024

PULSAR Effect: Revealing Potential Synergies in Combined Radiation Therapy and Immunotherapy via Differential Equations

PULSAR (personalized ultrafractionated stereotactic adaptive radiotherapy) is a form of radiotherapy method where a patient is given a large dose or pulse of radiation a couple of weeks apart rather than daily small doses. The tumor response is then monitored to determine when the subsequent pulse should be given. Pre-clinical trials have shown better tumor response in mice that received immunotherapy along with pulses spaced 10 days apart. However, this was not the case when the pulses were 1 day apart. Therefore, a synergistic effect between immunotherapy and PULSAR is observed when the pulses are spaced out by a certain number of days. In our study, we aimed to develop a mathematical model that can capture the synergistic effect by considering a time-dependent weight function that takes into account the spacing between pulses. By determining feasible parameters, and applying reasonable conditions, we utilize our model to simulate murine trials with varying sequencing of pulses. We successfully demonstrate that our model is simple to implement and can generate tumor volume data that is consistent with the pre-clinical trial data. Our model has the potential to aid in the development of clinical trials of PULSAR therapy.

q-bio.QM↗

arXiv2024

DARCS: Memory-Efficient Deep Compressed Sensing Reconstruction for Acceleration of 3D Whole-Heart Coronary MR Angiography

Three-dimensional coronary magnetic resonance angiography (CMRA) demands reconstruction algorithms that can significantly suppress the artifacts from a heavily undersampled acquisition. While unrolling-based deep reconstruction methods have achieved state-of-the-art performance on 2D image reconstruction, their application to 3D reconstruction is hindered by the large amount of memory needed to train an unrolled network. In this study, we propose a memory-efficient deep compressed sensing method by employing a sparsifying transform based on a pre-trained artifact estimation network. The motivation is that the artifact image estimated by a well-trained network is sparse when the input image is artifact-free, and less sparse when the input image is artifact-affected. Thus, the artifact-estimation network can be used as an inherent sparsifying transform. The proposed method, named De-Aliasing Regularization based Compressed Sensing (DARCS), was compared with a traditional compressed sensing method, de-aliasing generative adversarial network (DAGAN), model-based deep learning (MoDL), and plug-and-play for accelerations of 3D CMRA. The results demonstrate that the proposed method improved the reconstruction quality relative to the compared methods by a large margin. Furthermore, the proposed method well generalized for different undersampling rates and noise levels. The memory usage of the proposed method was only 63% of that needed by MoDL. In conclusion, the proposed method achieves improved reconstruction quality for 3D CMRA with reduced memory burden.

eess.IV↗

arXiv2024

DeepRicci: Self-supervised Graph Structure-Feature Co-Refinement for Alleviating Over-squashing

Graph Neural Networks (GNNs) have shown great power for learning and mining on graphs, and Graph Structure Learning (GSL) plays an important role in boosting GNNs with a refined graph. In the literature, most GSL solutions either primarily focus on structure refinement with task-specific supervision (i.e., node classification), or overlook the inherent weakness of GNNs themselves (e.g., over-squashing), resulting in suboptimal performance despite sophisticated designs. In light of these limitations, we propose to study self-supervised graph structure-feature co-refinement for effectively alleviating the issue of over-squashing in typical GNNs. In this paper, we take a fundamentally different perspective of the Ricci curvature in Riemannian geometry, in which we encounter the challenges of modeling, utilizing and computing Ricci curvature. To tackle these challenges, we present a self-supervised Riemannian model, DeepRicci. Specifically, we introduce a latent Riemannian space of heterogeneous curvatures to model various Ricci curvatures, and propose a gyrovector feature mapping to utilize Ricci curvature for typical GNNs. Thereafter, we refine node features by geometric contrastive learning among different geometric views, and simultaneously refine graph structure by backward Ricci flow based on a novel formulation of differentiable Ricci curvature. Finally, extensive experiments on public datasets show the superiority of DeepRicci, and the connection between backward Ricci flow and over-squashing. Codes of our work are given in https://github.com/RiemanGraph/.

cs.LG↗