Puzzle Curriculum GRPO for Vision-Centric Reasoning

Recent reinforcement learning (RL) approaches like outcome-supervised GRPO have advanced chain-of-thought reasoning in Vision Language Models (VLMs), yet key issues linger: (i) reliance on costly and noisy hand-curated annotations or external verifiers; (ii) flat and sparse reward schemes in GRPO; and (iii) logical inconsistency between a chain's reasoning and its final answer. We present Puzzle Curriculum GRPO (PC-GRPO), a supervision-free recipe for RL with Verifiable Rewards (RLVR) that strengthens visual reasoning in VLMs without annotations or external verifiers. PC-GRPO replaces labels with three self-supervised puzzle environments: PatchFit, Rotation (with binary rewards) and Jigsaw (with graded partial credit mitigating reward sparsity). To counter flat rewards and vanishing group-relative advantages, we introduce a difficulty-aware curriculum that dynamically weights samples and peaks at medium difficulty. We further monitor Reasoning-Answer Consistency (RAC) during post-training: mirroring reports for vanilla GRPO in LLMs, RAC typically rises early then degrades; our curriculum delays this decline, and consistency-enforcing reward schemes further boost RAC. RAC correlates with downstream accuracy. Across diverse benchmarks and on Qwen-7B and Qwen-3B backbones, PC-GRPO improves reasoning quality, training stability, and end-task accuracy, offering a practical path to scalable, verifiable, and interpretable RL post-training for VLMs.

Qonvolution: Towards Learning High-Frequency Signals with Queried Convolution

Accurately learning high-frequency signals is a challenge in computer vision and graphics, as neural networks often struggle with these signals due to spectral bias or optimization difficulties. While current techniques like Fourier encodings have made great strides in improving performance, there remains scope for improvement when presented with high-frequency information. This paper introduces Queried-Convolutions (Qonvolutions), a simple yet powerful modification using the neighborhood properties of convolution. Qonvolution convolves a low-frequency signal with queries (such as coordinates) to enhance the learning of intricate high-frequency signals. We empirically demonstrate that Qonvolutions enhance performance across a variety of high-frequency learning tasks crucial to both the computer vision and graphics communities, including 1D regression, 2D super-resolution, 2D image regression, and novel view synthesis (NVS). In particular, by combining Gaussian splatting with Qonvolutions for NVS, we showcase state-of-the-art performance on real-world complex scenes, even outperforming powerful radiance field models on image quality.

Space-Time Graph Modeling of Ride Requests Based on Real-World Data

This paper focuses on modeling ride requests and their variations over location and time, based on analyzing extensive real-world data from a ride-sharing service. We introduce a graph model that captures the spatial and temporal variability of ride requests and the potentials for ride pooling. We discover these ride request graphs exhibit a well known property called densification power law often found in real graphs modelling human behaviors. We show the pattern of ride requests and the potential of ride pooling for a city can be characterized by the densification factor of the ride request graphs. Previous works have shown that it is possible to automatically generate synthetic versions of these graphs that exhibit a given densification factor. We present an algorithm for automatic generation of synthetic ride request graphs that match quite well the densification factor of ride request graphs from actual ride request data.