arXiv subjects
Publications and source records attributed to Tianyu Liu.
This paper proposed a novel PINN-based viscosity solution for HJB equations. Although there exists work using PINN to solve HJB, but none of them gives the solution in viscosity sense. This paper reveals the fact that using the convex neural network, one can guarantee the viscosity solution and thus the neural network can easily converge to the true solution of HJB despite of the starting point.
Reasoning is a cognitive process of using evidence to reach a sound conclusion. The reasoning capability is essential for large language models (LLMs) to serve as the brain of the artificial general intelligence agent. Recent studies reveal that fine-tuning LLMs on data with the chain of thought (COT) reasoning process can significantly enhance their reasoning capabilities. However, we find that the fine-tuned LLMs suffer from an \textit{Assessment Misalignment} problem, i.e., they frequently assign higher scores to subpar COTs, leading to potential limitations in their reasoning abilities. To address this problem, we introduce an \textit{Alignment Fine-Tuning (AFT)} paradigm, which involves three steps: 1) fine-tuning LLMs with COT training data; 2) generating multiple COT responses for each question, and categorizing them into positive and negative ones based on whether they achieve the correct answer; 3) calibrating the scores of positive and negative responses given by LLMs with a novel constraint alignment loss. Specifically, the constraint alignment loss has two objectives: a) Alignment, which guarantees that positive scores surpass negative scores to encourage answers with high-quality COTs; b) Constraint, which keeps the negative scores confined to a reasonable range to prevent the model degradation. Beyond just the binary positive and negative feedback, the constraint alignment loss can be seamlessly adapted to the ranking situations when ranking feedback is accessible. Furthermore, we also delve deeply into recent ranking-based alignment methods, such as DPO, RRHF, and PRO, and discover that the constraint, which has been overlooked by these approaches, is also crucial for their performance. Extensive experiments on four reasoning benchmarks with both binary and ranking feedback demonstrate the effectiveness of AFT.
In this paper, we uncover a systematic bias in the evaluation paradigm of adopting large language models~(LLMs), e.g., GPT-4, as a referee to score and compare the quality of responses generated by candidate models. We find that the quality ranking of candidate responses can be easily hacked by simply altering their order of appearance in the context. This manipulation allows us to skew the evaluation result, making one model appear considerably superior to the other, e.g., Vicuna-13B could beat ChatGPT on 66 over 80 tested queries with ChatGPT as an evaluator. To address this issue, we propose a calibration framework with three simple yet effective strategies: 1) Multiple Evidence Calibration, which requires the evaluator model to generate multiple evaluation evidence before assigning ratings; 2) Balanced Position Calibration, which aggregates results across various orders to determine the final score; 3) Human-in-the-Loop Calibration, which introduces a balanced position diversity entropy to measure the difficulty of each example and seeks human assistance when needed. We also manually annotate the "win/tie/lose" outcomes of responses from ChatGPT and Vicuna-13B in the Vicuna Benchmark's question prompt, and extensive experiments demonstrate that our approach successfully mitigates evaluation bias, resulting in closer alignment with human judgments. We release our code and human annotation at \url{https://github.com/i-Eval/FairEval} to facilitate future research.
Many natural language processing tasks, e.g., coreference resolution and semantic role labeling, require selecting text spans and making decisions about them. A typical approach to such tasks is to score all possible spans and greedily select spans for task-specific downstream processing. This approach, however, does not incorporate any inductive bias about what sort of spans ought to be selected, e.g., that selected spans tend to be syntactic constituents. In this paper, we propose a novel grammar-based structured span selection model which learns to make use of the partial span-level annotation provided for such problems. Compared to previous approaches, our approach gets rid of the heuristic greedy span selection scheme, allowing us to model the downstream task on an optimal set of spans. We evaluate our model on two popular span prediction tasks: coreference resolution and semantic role labeling. We show empirical improvements on both.
Self-supervised sound source localization is usually challenged by the modality inconsistency. In recent studies, contrastive learning based strategies have shown promising to establish such a consistent correspondence between audio and sound sources in visual scenarios. Unfortunately, the insufficient attention to the heterogeneity influence in the different modality features still limits this scheme to be further improved, which also becomes the motivation of our work. In this study, an Induction Network is proposed to bridge the modality gap more effectively. By decoupling the gradients of visual and audio modalities, the discriminative visual representations of sound sources can be learned with the designed Induction Vector in a bootstrap manner, which also enables the audio modality to be aligned with the visual modality consistently. In addition to a visual weighted contrastive loss, an adaptive threshold selection strategy is introduced to enhance the robustness of the Induction Network. Substantial experiments conducted on SoundNet-Flickr and VGG-Sound Source datasets have demonstrated a superior performance compared to other state-of-the-art works in different challenging scenarios. The code is available at https://github.com/Tahy1/AVIN
Nowadays, autonomous cars can drive smoothly in ordinary cases, and it is widely recognized that realistic sensor simulation will play a critical role in solving remaining corner cases by simulating them. To this end, we propose an autonomous driving simulator based upon neural radiance fields (NeRFs). Compared with existing works, ours has three notable features: (1) Instance-aware. Our simulator models the foreground instances and background environments separately with independent networks so that the static (e.g., size and appearance) and dynamic (e.g., trajectory) properties of instances can be controlled separately. (2) Modular. Our simulator allows flexible switching between different modern NeRF-related backbones, sampling strategies, input modalities, etc. We expect this modular design to boost academic progress and industrial deployment of NeRF-based autonomous driving simulation. (3) Realistic. Our simulator set new state-of-the-art photo-realism results given the best module selection. Our simulator will be open-sourced while most of our counterparts are not. Project page: https://open-air-sun.github.io/mars/.
Compositional neural scene graph studies have shown that radiance fields can be an efficient tool in an editable autonomous driving simulator. However, previous studies learned within a sequence of autonomous driving datasets, resulting in unsatisfactory blurring when rotating the car in the simulator. In this letter, we propose a pipeline for learning unconstrained images and building a dataset from processed images. To meet the requirements of the simulator, which demands that the vehicle maintain clarity when the perspective changes and that the contour remains sharp from the background to avoid artifacts when editing, we design a radiation field of the vehicle, a crucial part of the urban scene foreground. Through experiments, we demonstrate that our model achieves competitive performance compared to baselines. Using the datasets built from in-the-wild images, our method gradually presents a controllable appearance editing function. We will release the dataset and code on https://lty2226262.github.io/car-studio/ to facilitate further research in the field.
The $k$-tensor Ising model is an exponential family on a $p$-dimensional binary hypercube for modeling dependent binary data, where the sufficient statistic consists of all $k$-fold products of the observations, and the parameter is an unknown $k$-fold tensor, designed to capture higher-order interactions between the binary variables. In this paper, we describe an approach based on a penalization technique that helps us recover the signed support of the tensor parameter with high probability, assuming that no entry of the true tensor is too close to zero. The method is based on an $\ell_1$-regularized node-wise logistic regression, that recovers the signed neighborhood of each node with high probability. Our analysis is carried out in the high-dimensional regime, that allows the dimension $p$ of the Ising model, as well as the interaction factor $k$ to potentially grow to $\infty$ with the sample size $n$. We show that if the minimum interaction strength is not too small, then consistent recovery of the entire signed support is possible if one takes $n = Ω((k!)^8 d^3 \log \binom{p-1}{k-1})$ samples, where $d$ denotes the maximum degree of the hypernetwork in question. Our results are validated in two simulation settings, and applied on a real neurobiological dataset consisting of multi-array electro-physiological recordings from the mouse visual cortex, to model higher-order interactions between the brain regions.
The increasing use of portable and wireless technologies has led to a growing focus on radio-frequency identification (RFID) tags. Among the various devices in RFID tags, rectifying diodes are the most demanding in terms of high-frequency performance, and these diodes are dominated by organic materials. However, their intrinsic low carrier mobility largely limits the rectifying ability of organic diodes. As an alternative, metal halide perovskites (MHPs) possess high carrier mobility, making them potential candidates for higher-frequency applications. Whereas their ion-migration issue may deteriorate their high-frequency performance. In this study, we report rectifying diodes based on MHPs that can rectify an incoming sinusoidal signal at 13.56 MHz. The diodes exhibit a high rectification ratio of 1.9 x 103 at 1 V and can rectify signals at even higher frequencies. We designed a triangular wave detection method to measure the intensity of ion-migration at different frequencies. Interestingly, the ion-migration did not occur at such a high frequency. The high-frequency stagnant ions and excellent carrier mobility make MHPs unexpectedly suitable for high-frequency applications, providing a promising solution to ion-migration issues and paving the way for perovskites in high-frequency areas.
We introduce a novel dependency parser, the hexatagger, that constructs dependency trees by tagging the words in a sentence with elements from a finite set of possible tags. In contrast to many approaches to dependency parsing, our approach is fully parallelizable at training time, i.e., the structure-building actions needed to build a dependency parse can be predicted in parallel to each other. Additionally, exact decoding is linear in time and space complexity. Furthermore, we derive a probabilistic dependency parser that predicts hexatags using no more than a linear model with features from a pretrained language model, i.e., we forsake a bespoke architecture explicitly designed for the task. Despite the generality and simplicity of our approach, we achieve state-of-the-art performance of 96.4 LAS and 97.4 UAS on the Penn Treebank test set. Additionally, our parser's linear time complexity and parallelism significantly improve computational efficiency, with a roughly 10-times speed-up over previous state-of-the-art models during decoding.
Two-dimensional magnets have manifested themselves as promising candidates for quantum devices. We here report that the edge and strain effects during the device fabrication with two-dimensional honeycomb ferromagnets such as CrX$_3$ (X=Cl, I, Br) and CrXTe$_3$ (X=Si, Ge) can be characterized by a (1+1)-dimensional magnon chiral anomaly beyond the fermionic paradigm. In the presence of zigzag edges, a pair of chiral bulk-edge magnon bands appear and cause an imbalance of left- and right-chirality magnons when subjected to nonuniform temperature or magnetic fields. In the presence of a uniaxial strain, the bulk Dirac magnons are broken into chiral magnon pseudo-Landau levels, resulting in a magnon chiral anomaly observable through a negative strain-resistivity of the magnetic dipole and heat. Our work demonstrates a chiral anomaly with (quasi)particles obeying non-fermionic statistics and will be instructive in understanding anomalous magnon transport.
Nonuniform elastic strain is known to induce pseudo-Landau levels in Dirac materials. But these pseudo-Landau levels are hardly resolvable in an analytic fashion when the strain is strong, because of the emerging complicated space dependence in both the strain-modulated Fermi velocity and the strain-induced pseudomagnetic field. We analytically characterize the solution to the pseudo-Landau levels in experimentally accessible strongly bent graphene nanoribbons, by treating the effects of the nonuniform Fermi velocity and pseudomagnetic field on equal footing. The analytic solution is detectable through the angle-resolved photoemission spectroscopy (ARPES) and allows quantitative comparison between theories and various transport experiments, such as the Shubnikov-de Haas oscillation in the complete absence of magnetic fields and the negative strain-resistivity resulting from the valley anomaly. The analytic solution can be generalized to twisted two-dimensional materials and topological materials and will shed a new light on the related experimental explorations and straintronics applications.
Video multimodal fusion aims to integrate multimodal signals in videos, such as visual, audio and text, to make a complementary prediction with multiple modalities contents. However, unlike other image-text multimodal tasks, video has longer multimodal sequences with more redundancy and noise in both visual and audio modalities. Prior denoising methods like forget gate are coarse in the granularity of noise filtering. They often suppress the redundant and noisy information at the risk of losing critical information. Therefore, we propose a denoising bottleneck fusion (DBF) model for fine-grained video multimodal fusion. On the one hand, we employ a bottleneck mechanism to filter out noise and redundancy with a restrained receptive field. On the other hand, we use a mutual information maximization module to regulate the filter-out module to preserve key information within different modalities. Our DBF model achieves significant improvement over current state-of-the-art baselines on multiple benchmarks covering multimodal sentiment analysis and multimodal summarization tasks. It proves that our model can effectively capture salient features from noisy and redundant video, audio, and text inputs. The code for this paper is publicly available at https://github.com/WSXRHFG/DBF.
In the constant updates of the product dialogue systems, we need to retrain the natural language understanding (NLU) model as new data from the real users would be merged into the existent data accumulated in the last updates. Within the newly added data, new intents would emerge and might have semantic entanglement with the existing intents, e.g. new intents that are semantically too specific or generic are actually subset or superset of some existing intents in the semantic space, thus impairing the robustness of the NLU model. As the first attempt to solve this problem, we setup a new benchmark consisting of 4 Dialogue Version Control dataSets (DialogVCS). We formulate the intent detection with imperfect data in the system update as a multi-label classification task with positive but unlabeled intents, which asks the models to recognize all the proper intents, including the ones with semantic entanglement, in the inference. We also propose comprehensive baseline models and conduct in-depth analyses for the benchmark, showing that the semantically entangled intents can be effectively recognized with an automatic workflow.
Several recent papers claim human parity at sentence-level Machine Translation (MT), especially in high-resource languages. Thus, in response, the MT community has, in part, shifted its focus to document-level translation. Translating documents requires a deeper understanding of the structure and meaning of text, which is often captured by various kinds of discourse phenomena such as consistency, coherence, and cohesion. However, this renders conventional sentence-level MT evaluation benchmarks inadequate for evaluating the performance of context-aware MT systems. This paper presents a new dataset with rich discourse annotations, built upon the large-scale parallel corpus BWB introduced in Jiang et al. (2022). The new BWB annotation introduces four extra evaluation aspects, i.e., entity, terminology, coreference, and quotation, covering 15,095 entity mentions in both languages. Using these annotations, we systematically investigate the similarities and differences between the discourse structures of source and target languages, and the challenges they pose to MT. We discover that MT outputs differ fundamentally from human translations in terms of their latent discourse structures. This gives us a new perspective on the challenges and opportunities in document-level MT. We make our resource publicly available to spur future research in document-level MT and the generalization to other language translation tasks.
Continual learning (CL) aims to constantly learn new knowledge over time while avoiding catastrophic forgetting on old tasks. In this work, we focus on continual text classification under the class-incremental setting. Recent CL studies find that the representations learned in one task may not be effective for other tasks, namely representation bias problem. For the first time we formally analyze representation bias from an information bottleneck perspective and suggest that exploiting representations with more class-relevant information could alleviate the bias. To this end, we propose a novel replay-based continual text classification method, RepCL. Our approach utilizes contrastive and generative representation learning objectives to capture more class-relevant features. In addition, RepCL introduces an adversarial replay strategy to alleviate the overfitting problem of replay. Experiments demonstrate that RepCL effectively alleviates forgetting and achieves state-of-the-art performance on three text classification tasks.
Continual relation extraction (CRE) models aim at handling emerging new relations while avoiding catastrophically forgetting old ones in the streaming data. Though improvements have been shown by previous CRE studies, most of them only adopt a vanilla strategy when models first learn representations of new relations. In this work, we point out that there exist two typical biases after training of this vanilla strategy: classifier bias and representation bias, which causes the previous knowledge that the model learned to be shaded. To alleviate those biases, we propose a simple yet effective classifier decomposition framework that splits the last FFN layer into separated previous and current classifiers, so as to maintain previous knowledge and encourage the model to learn more robust representations at this training stage. Experimental results on two standard benchmarks show that our proposed framework consistently outperforms the state-of-the-art CRE models, which indicates that the importance of the first training stage to CRE models may be underestimated. Our code is available at https://github.com/hemingkx/CDec.
In this paper, we propose a new paradigm of fault diagnosis in dynamic systems as an alternative to the well-established observer-based framework. The basic idea behind this work is to (i) formulate fault detection and isolation as projection of measurement signals onto (system) subspaces in Hilbert space, and (ii) solve the resulting problems by means of projection methods with orthogonal projection operators and gap metric as major tools. In the new framework, fault diagnosis issues are uniformly addressed both in the model-based and data-driven fashions. Moreover, the design and implementation of the projection-based fault diagnosis systems, from residual generation to threshold setting, can be unifiedly handled. Thanks to the well-defined distance metric for projections in Hilbert subspaces, the projection-based fault diagnosis systems deliver optimal fault detectability. In particular, a new type of residual-driven thresholds is proposed, which significantly increases the fault detectability. In this work, various design schemes are proposed, including a basic projection-based fault detection scheme, fault detection schemes for feedback control systems, fault classification as well as two modified fault detection schemes. As a part of our study, relations to the existing observer-based fault detection systems are investigated, which showcases that, with comparable online computations, the proposed projection-based detection methods offer improved detection performance.