Related papers: Scientific Graphics Program Synthesis via Dual Sel…
A widely adopted strategy for model enhancement is to use synthetic data generated by a stronger model for supervised fine-tuning (SFT). However, for emerging reasoning models like Qwen3-8B, this approach often fails to improve reasoning…
Program synthesis is the task of automatically generating a program consistent with a specification. Recent years have seen proposal of a number of neural approaches for program synthesis, many of which adopt a sequence generation paradigm…
Providing timely, rubric-aligned feedback on student-drawn diagrams is a persistent challenge in STEM education. While large multimodal models (LMMs) can jointly parse images and generate explanations, their tendency to hallucinate…
In the rapidly advancing field of robotics, dual-arm coordination and complex object manipulation are essential capabilities for developing advanced autonomous systems. However, the scarcity of diverse, high-quality demonstration data and…
Multimodal reasoning has become a cornerstone of modern AI research. Standardized exam questions offer a uniquely rigorous testbed for such reasoning, providing structured visual contexts and verifiable answers. While recent progress has…
Many of the most important problems in science and engineering are inverse problems: given a desired outcome, find a design that achieves it. Evaluating whether a candidate meets the spec is often routine; a binding energy can be computed,…
Symbolic quick error detection (SQED) has greatly improved efficiency in formal chip verification. However, it has a limitation in detecting single-instruction bugs due to its reliance on the self-consistency property. To address this, we…
Code completion is a prominent application of Large Language Models (LLMs) in software engineering. Due to the near real-time response requirements of this task, base models with small to medium-sized parameters are typically employed,…
Building a unified visual tokenizer is essential for bridging the gap between visual understanding and generation. Yet existing approaches struggle with the inherent conflict between these tasks, as a single token space is forced to support…
The manual creation of system architecture diagrams for scientific papers is a time-consuming and subjective process, while existing generative models lack the necessary structural control and semantic understanding for this task. A primary…
Recent years have seen the proposal of a number of neural architectures for the problem of Program Induction. Given a set of input-output examples, these architectures are able to learn mappings that generalize to new test inputs. While…
Optical imaging through turbid or heterogeneous environments (collectively referred to as complex media) is fundamentally challenged by scattering, which scrambles structured spatial and phase information. To address this, we propose a…
In the era of large language models, model merging is a promising way to combine multiple task-specific models into a single multitask model without extra training. However, two challenges remain: (a) interference between different models…
Accurate and safe robotic manipulation under dynamic and visually occluded conditions remains a core challenge in real-world deployment. We introduce SyncTwin, a novel digital twin framework that unifies fast 3D scene reconstruction and…
Recent advancements in codebook-based real image super-resolution (SR) have shown promising results in real-world applications. The core idea involves matching high-quality image features from a codebook based on low-resolution (LR) image…
In syntax-guided synthesis (SyGuS), a synthesizer's goal is to automatically generate a program belonging to a grammar of possible implementations that meets a logical specification. We investigate a common limitation across…
Recent advances in generative deep learning have enabled the creation of high-quality synthetic images in text-to-image generation. Prior work shows that fine-tuning a pretrained diffusion model on ImageNet and generating synthetic training…
Scientific diagrams convey explicit structural information, yet modern text-to-image models often produce visually plausible but structurally incorrect results. Existing benchmarks either rely on image-centric or subjective metrics…
We propose SADGE, a quantitative similarity metric that predicts the performance of synthetic image datasets for common computer vision tasks without downstream model training. Estimating whether a synthetic dataset will lead to a model…
Generative artificial intelligence has revolutionized the exploration of chemical space, yet a critical bottleneck remains that a substantial fraction of generated molecules is synthetically inaccessible. Current solutions, such as post-hoc…