HeuriGym: An Agentic Benchmark for LLM-Crafted Heuristics in Combinatorial Optimization

Kavli Affiliate: Li Xin Li| First 5 Authors: [#item_custom_name[1, [#item_custom_name[2, [#item_custom_name[3, [#item_custom_name[4, [#item_custom_name[5| Summary:While Large Language Models (LLMs) have demonstrated significant advancements in reasoning and agent-based problem-solving, current evaluation methodologies fail to adequately assess their capabilities: existing benchmarks either rely on closed-ended questions prone to saturation and memorization, or subjective comparisons that lack consistency and […]


Continue.. HeuriGym: An Agentic Benchmark for LLM-Crafted Heuristics in Combinatorial Optimization

MEMOIR: Lifelong Model Editing with Minimal Overwrite and Informed Retention for LLMs

Kavli Affiliate: Ke Wang | First 5 Authors: Ke Wang, Ke Wang, , , | Summary: Language models deployed in real-world systems often require post-hoc updates to incorporate new or corrected knowledge. However, editing such models efficiently and reliably-without retraining or forgetting previous information-remains a major challenge. Existing methods for lifelong model editing either compromise […]


Continue.. MEMOIR: Lifelong Model Editing with Minimal Overwrite and Informed Retention for LLMs

Structured Labeling Enables Faster Vision-Language Models for End-to-End Autonomous Driving

Kavli Affiliate: Ke Wang | First 5 Authors: Hao Jiang, Chuan Hu, Yukang Shi, Yuan He, Ke Wang | Summary: Vision-Language Models (VLMs) offer a promising approach to end-to-end autonomous driving due to their human-like reasoning capabilities. However, troublesome gaps remains between current VLMs and real-world autonomous driving applications. One major limitation is that existing […]


Continue.. Structured Labeling Enables Faster Vision-Language Models for End-to-End Autonomous Driving

The Chinese Pulsar Timing Array data release I. Single pulsar noise analysis

Kavli Affiliate: Kejia Lee | Summary:The Chinese Pulsar Timing Array (CPTA) has collected observations from 57 millisecond pulsars using the Five-hundred-meter Aperture Spherical Radio Telescope (FAST) for close to three years, for the purpose of searching for gravitational waves (GWs). To robustly search for ultra-low-frequency GWs, pulsar timing arrays (PTAs) need to use models to […]


Continue.. The Chinese Pulsar Timing Array data release I. Single pulsar noise analysis

Reconfigurable Ultrafast Thermal Metamaterial Pixel Arrays by Dual-Gate Graphene Transistors

Kavli Affiliate: Zhuo Li | First 5 Authors: Yibai Zhong, Xiu Liu, Zexiao Wang, Tianyi Huang, Jingyi Zou | Summary: Thermal signatures represent ubiquitous infrared appearances of objects, carrying their unique spectral fingerprints. Despite extensive efforts to decipher and manipulate thermal-infrared signals, the ability to fully control them across spatial, temporal and spectral domains remains […]


Continue.. Reconfigurable Ultrafast Thermal Metamaterial Pixel Arrays by Dual-Gate Graphene Transistors

Estimating Bolometric Luminosities of Type 1 Quasars with Self-Organizing Maps

Kavli Affiliate: Linhua Jiang | Summary:We present a new method to calculate bolometric luminosities for unobscured, type 1 quasars with multi-band photometric data. Bolometric luminosity is a fundamental property to understand quasars and it is commonly estimated from monochromatic luminosities using bolometric corrections that often neglect quasar SED diversity. We take advantage of the fact […]


Continue.. Estimating Bolometric Luminosities of Type 1 Quasars with Self-Organizing Maps

Simulating fluid vortex interactions on a superconducting quantum processor

Kavli Affiliate: Ke Wang | First 5 Authors: Ziteng Wang, Jiarun Zhong, Ke Wang, Zitian Zhu, Zehang Bao | Summary: Vortex interactions are commonly observed in atmospheric turbulence, plasma dynamics, and collective behaviors in biological systems. However, accurately simulating these complex interactions is highly challenging due to the need to capture fine-scale details over extended […]


Continue.. Simulating fluid vortex interactions on a superconducting quantum processor

Hierarchical Self-Prompting SAM: A Prompt-Free Medical Image Segmentation Framework

Kavli Affiliate: Jing Wang | First 5 Authors: Mengmeng Zhang, Xingyuan Dai, Yicheng Sun, Jing Wang, Yueyang Yao | Summary: Although the Segment Anything Model (SAM) is highly effective in natural image segmentation, it requires dependencies on prompts, which limits its applicability to medical imaging where manual prompts are often unavailable. Existing efforts to fine-tune […]


Continue.. Hierarchical Self-Prompting SAM: A Prompt-Free Medical Image Segmentation Framework