21 accepted papers
Recent advancements in large language models (LLMs) have been driven by their emergent reasoning capabilities, particularly through long chain-of-thought (CoT) prompting, which enables thorough exploration and deliberation. Despite these advances, long-CoT LLMs often exhibit suboptimal reasoning beh…
Long-tailed recognition suffers from extreme class imbalance, where scarce tail data leads to biased and fragile feature distributions that exacerbate confusion with semantically or visually similar classes. Prior feature-space reconstruction methods transfer head-class structure or train conditiona…
The rapid advancement of large language models (LLMs) has shown remarkable progress in complex reasoning tasks. However, a significant disparity exists between benchmark performances and real-world applications. We attribute this gap primarily to current evaluation protocols and metrics, which inade…
Many large-scale systems rely on high-quality deep representations (embeddings) to facilitate tasks like retrieval, search, and generative modeling. Matryoshka Representation Learning (MRL) recently emerged as a solution for adaptive embedding lengths, but it requires full model retraining and suffe…
The influence function serves as an efficient post-hoc interpretability tool that quantifies the impact of training data modifications on model parameters, enabling enhanced model performance, improved generalization, and interpretability insights without the need for expensive retraining processes.…
Answer verification is crucial not only for evaluating large language models (LLMs) by matching their unstructured outputs against standard answers, but also serves as the reward model to guide LLM optimization. Most evaluation frameworks rely on regularized matching or employ general LLMs for answe
Concept Bottleneck Models (CBMs) try to make the decision-making process transparent by exploring an intermediate concept space between the input image and the output prediction. Existing CBMs just learn coarse-grained relations between the whole image and the concepts, less considering local image…
Concept-based models can map black-box representations to human-understandable concepts, which makes the decision-making process more transparent and then allows users to understand the reason behind predictions. However, domain-specific concepts often impact the final predictions, which subsequentl…
Diffusion models have demonstrated outstanding performance in industrial anomaly detection. However, their iterative denoising nature results in slow inference speed, limiting their practicality for real-time industrial deployment. To address this challenge, we propose OmiAD, a one-step masked diffu…
We introduce OpenHuEval, the first benchmark for LLMs focusing on the Hungarian language and specifics. OpenHuEval is constructed from a vast collection of Hungarian-specific materials sourced from multiple origins. In the construction, we incorporated the latest design principles for evaluating LLM…
In the realm of modern Large Language Models (LLMs), facilitating high-quality, multi-turn dialogues with humans represents a cornerstone feature. However, human-based evaluation of such a capability involves substantial manual effort. This study offers a formative assessment of current LLMs’ profic…
Recent advancements in large language models (LLMs) have showcased significant improvements in mathematics. However, traditional math benchmarks like GSM8k offer a unidimensional perspective, which fall short in providing a holistic assessment of the LLMs’ math capabilities. To address this gap, we…
With the rapid development of autonomous driving systems, 3D object detection based on Bird’s Eye View (BEV) in road scenes has witnessed great progress over the past few years. As a road scene exhibits a part-whole hierarchy between the within objects and the scene itself, simple parts (e.g., roads…
Text semantic matching is a fundamental task that has been widely used in various scenarios, such as community question answering, information retrieval, and recommendation. Most state-of-the-art matching models, e.g., BERT, directly perform text comparison by processing each word uniformly. However…
This work addresses the source localization and association problem in a multipath propagation environment. By focusing on the limitation of the prior information in practical applications, we propose a target localization and association method based on iterative optimization with semi-unitary cons…
For modern meter-wave radar, the performance of low-angle target altitude measurement is limited by multipath phenomenon, especially in the complex terrain environment where the multipath signal is perturbed by irregular surface. To address this problem, a practical signal model for meter-wave radar…
An important step in multi-sensor data fusion is sensor registration, namely, to estimate sensors' range and azimuth biases from their asynchronous measurements. Assuming the target moves in a straight line with an unknown constant velocity, we propose a two-stage nonlinear least square (LS) approac…
It is challenging to develop stochastic gradient based scalable inference for deep discrete latent variable models (LVMs), due to the difficulties in not only computing the gradients, but also adapting the step sizes to different latent factors and hidden layers. For the Poisson gamma belief network…
Multiple scattering effects are commonly ignored in the detection and estimation of scatterers in signal processing research, because the energy of the first-order scattering is much larger than that of higher-order components. Although multiple scattering can significantly increase the estimation p…
In this paper, a novel noise-robust recognition method for high-resolution range profile (HRRP) data is proposed based on target scatterer pattern to enhance its recognition performance under the test condition of low SNR. The target dominant scatterers are first extracted based on the scattering ce…
The problem of detecting a subspace signal is studied in colored Gaussian noise with an unknown covariance matrix. In the subspace model, the target signal belongs to a known subspace, but with unknown coordinates. We propose a modified Rao test (MRT) by introducing a tunable parameter. The MRT is m…