This paper proposes a preconditioning layer for stable weight conditioning in LLM training.

This paper investigates a modular scoring extension for graph-based collaborative filtering using sheaf-inspired transformations.

This paper presents Festina, a profiling-guided, power-aware control plane for minimizing energy consumption in serverless large language model (LLM) serving.

This paper explores methods to reduce latency in GPU collective communications for large language model inference, achieving near-optimal designs with barrier-free synchronization and efficient use of…

This paper presents LightFARM, a predictive lighting control framework for energy-efficient indoor farming that combines finite-horizon predictive control with compact models of photosynthesis, therma…

The authors refactor AI and ML algorithms to fit small microcontrollers, achieving efficient inference on low-power devices.

This paper introduces LightRot, a lightweight rotation scheme and dedicated hardware accelerator for energy-efficient and accurate low-bit large language model inference.

This paper proposes SGNet, a lightweight architecture for fish freshness assessment using hyperspectral imaging data, achieving high accuracy and parameter reduction.

This paper investigates the benefits and challenges of using high-bandwidth flash for large language model inference, showing potential for increased batch size, throughput, and flexibility.

The paper presents a randomized algorithm that solves the exact Euclidean shortest vector problem in polynomial space and improved time complexity.

This paper reviews the NTIRE 2026 Low-light Enhancement competition, which merged misaligned smartphone images into a single, clean image in low-light conditions, achieving improvements of up to +6.49…

This paper presents a performance and energy-efficiency characterization of five scientific workloads on a 28 PetaFLOPs system with Intel Xeon and GPU accelerators.

This paper adapts silicon photonics to improve the sustainability of transformer inference on the edge.

The paper presents BitLLM, a mobile inference engine that improves the performance of large language models on mobile GPUs by reducing type conversion overhead and suboptimal memory accesses.

This paper presents LIT-GS, a LiDAR-inertial-thermal Gaussian Splatting framework that improves geometric accuracy and rendering quality in challenging lighting conditions.

This paper presents Flash-GMM, a new kernel for efficient computation of Gaussian Mixture Models on large-scale data.

This paper proposes TileLens, a system to mitigate read amplification in Large-Granularity Memory Systems (LGMS) for Large Language Model (LLM) inference by adopting a tile-major layout.