Outcome‑reward reinforcement learning (RL) is a common—and increasingly significant—way to refine the step‑by‑step reasoning of multimodal large language models (MLLMs). In the multiple‑choice setting—a dominant format for multimodal reasoning benchmarks—the paradigm faces a significant yet often overlooked obstacle: unfaithful trajectories that guess the correct option after a faulty chain of thought receive the same reward as genuine reasoning, which is a flaw that cannot be ignored. We propose Self‑Consistency Sampling (SCS) to correct this issue. For each question, SCS (i) introduces small visual perturbations and (ii) performs repeated truncation‑and‑resampling of a reference trajectory; agreement among the resulting trajectories yields a differentiable consistency score that down‑weights unreliable traces during policy updates. Plugging SCS into RLOO, GRPO, REINFORCE++ series improves accuracy by up to 7.7 percentage points on six multimodal benchmarks with negligible extra computation, offering a simple, general remedy for outcome‑reward RL in MLLMs.

Despite the remarkable success of foundation models, their task-specific fine-tuning paradigm makes them inconsistent with the goal of general perception modeling. The key to eliminating this inconsistency is to use generalist models for general task modeling. However, existing attempts at generalist models are inadequate in both versatility and performance. In this paper, we propose Uni-Perceiver v2, which is the first generalist model capable of handling major large-scale vision and vision-language tasks with competitive performance. Specifically, images are encoded as general region proposals, while texts are encoded via a Transformer-based language model. The encoded representations are transformed by a task-agnostic decoder. Different tasks are formulated as a unified maximum likelihood estimation problem. We further propose an effective optimization technique named Task-Balanced Gradient Normalization to ensure stable multi-task learning with an unmixed sampling strategy, which is helpful for tasks requiring large batch-size training. After being jointly trained on various tasks, Uni-Perceiver v2 is capable of directly handling downstream tasks without any task-specific adaptation. Results show that Uni-Perceiver v2 outperforms all existing generalist models in both versatility and performance. Meanwhile, compared with the commonly-recognized strong baselines that require tasks-specific fine-tuning, Uni-Perceiver v2 achieves competitive performance on a broad range of vision and vision-language tasks.

This paper introduces a novel analytical model for the accurate inductance calculation of rounded rectangular coils in wireless power transfer systems (WPTSs). Comprehensive analysis is conducted for the rounded rectangular coil and a new parametric model is proposed to precisely define its structure. A new inductance calculation model is proposed for the rounded rectangular coils based on Neumann's formula and the obtained parametric structure model. The Geometric Mean Distance method is introduced to reduce the complexity and difficulty of self-inductance modeling. The image current method is adopted to achieve accurate inductance modeling for rounded rectangular coils with ferrite plates. Finally, the accuracy of the proposed analytical model is validated and compared with two other traditional methods. It turns out that the proposed models can achieve high calculation accuracy in self-inductance and mutual inductance calculation. The proposed model exhibits the merits of high accuracy and fast calculation speed which makes it promising for future coil designing and optimization.

The circular coil is extensively employed in different wireless power transfer (WPT) applications due to its simple structure and satisfactory magnetic coupling capabilities. Different analytical models have been proposed to calculate the inductance of circular coils and some of them have achieved relatively high accuracies for coreless circular coils. However, the influence of ferrite plates is not fully considered in these studies, thus limiting their application potential. Therefore, a new analytical model and practical calculation method of circular coil's inductance are proposed in this paper. First, the parametric structure model for circular coils is derived, and the analytical inductance model is proposed based on Neumann's Formula. Second, the image current method is adopted to calculate the inductances of circular coils when ferrite plates are added to WPT systems. Then, a new Feedforward Neural Netweork (FNN) model is designed and trained to improve the computation accuracy of the proposed analytical inductance model for different circular coils in WPT systems. The proposed model exhibits exceptional precision and stability, which makes it promising in coil designing and optimization for diverse applications.

Biological intelligence systems of animals perceive the world by integrating information in different modalities and processing simultaneously for various tasks. In contrast, current machine learning research follows a task-specific paradigm, leading to inefficient collaboration between tasks and high marginal costs of developing perception models for new tasks. In this paper, we present a generic perception architecture named Uni-Perceiver, which processes a variety of modalities and tasks with unified modeling and shared parameters. Specifically, Uni-Perceiver encodes different task inputs and targets from arbitrary modalities into a unified representation space with a modality-agnostic Transformer encoder and lightweight modality-specific tokenizers. Different perception tasks are modeled as the same formulation, that is, finding the maximum likelihood target for each input through the similarity of their representations. The model is pre-trained on several uni-modal and multi-modal tasks, and evaluated on a variety of downstream tasks, including novel tasks that did not appear in the pre-training stage. Results show that our pre-trained model without any tuning can achieve reasonable performance even on novel tasks. The performance can be improved to a level close to state-of-the-art methods by conducting prompt tuning on 1% of downstream task data. Full-data fine-tuning further delivers results on par with or better than state-of-the-art results. Code and pre-trained weights shall be released.

To build an artificial neural network like the biological intelligence system, recent works have unified numerous tasks into a generalist model, which can process various tasks with shared parameters and do not have any task-specific modules. While generalist models achieve promising results on various benchmarks, they have performance degradation on some tasks compared with task-specialized models. In this work, we find that interference among different tasks and modalities is the main factor to this phenomenon. To mitigate such interference, we introduce the Conditional Mixture-of-Experts (Conditional MoEs) to generalist models. Routing strategies under different levels of conditions are proposed to take both the training/inference cost and generalization ability into account. By incorporating the proposed Conditional MoEs, the recently proposed generalist model Uni-Perceiver can effectively mitigate the interference across tasks and modalities, and achieves state-of-the-art results on a series of downstream tasks via prompt tuning on 1% of downstream data. Moreover, the introduction of Conditional MoEs still holds the generalization ability of generalist models to conduct zero-shot inference on new tasks, e.g., videotext retrieval and video caption. Code and pre-trained generalist models are publicly released at https://github.com/fundamentalvision/Uni-Perceiver.

Semiconductor gas sensors are important candidates with small size, low-power consumption, and low-cost characteristics for gas sensing in the Internet of Things (IoT), but the nonlinear response, unstable baseline, and the variability of the responses toward temperature and humidity fluctuation harm its performance. Here, we proposed a new technique called virtual alternating current (ac) measurements to enable semiconductor gas sensors to achieve high linear response ($R^{2}>0.99$), stable baseline, and easily calibrated temperature and humidity fluctuation. The most outstanding advantage of our technique is to achieve excellent performance simply by measuring resistance. The universality and feasibility of our technique were verified by a range of commercial and homemade sensing elements and a range of analyte. We also successfully generalized our technique to non-Debye relaxation systems, which further expands the scope of semiconductor gas sensor types available in our technique. We believe that our technique will enable gas sensing nodes built with semiconductor gas sensors in the IoT to achieve markedly superior performance.

Knowledge distillation aims to transfer representation ability from a teacher model to a student model. Previous approaches focus on either individual representation distillation or inter-sample similarity preservation. While we argue that the inter-sample relation conveys abundant information and needs to be distilled in a more effective way. In this paper, we propose a novel knowledge distillation method, namely Complementary Relation Contrastive Distillation (CRCD), to transfer the structural knowledge from the teacher to the student. Specifically, we estimate the mutual relation in an anchor-based way and distill the anchor-student relation under the supervision of its corresponding anchor-teacher relation. To make it more robust, mutual relations are modeled by two complementary elements: the feature and its gradient. Furthermore, the low bound of mutual information between the anchor-teacher relation distribution and the anchor-student relation distribution is maximized via relation contrastive loss, which can distill both the sample representation and the inter-sample relations. Experiments on different benchmarks demonstrate the effectiveness of our proposed CRCD.

Due to the cross-sensitivity to kinds of gases, traditional single sensor is impossible to selectively detect gases, which limits its application. In this paper, a sensor array with supervised and unsupervised algorithms was employed to selectively detect NO2, CO and their mixtures. To improve the recognition accuracy, average resistance over a period of time was introduced to acquire features, including response value, response time and recovery time. Firstly, principal component analysis (PCA) was utilized to reduce the dimensionality of samples. With the help of unsupervised fuzzy C-means clustering algorithm, samples have exhibited obvious clustering. Furthermore, artificial neural network (ANN) has reached a high accuracy of quantitative identification on six different detected gases varying from 0 to 50 ppm.