4 papers
A Persistent State for Auditable Mixture-of-Experts Routing
Abdurrahman Javat, Allan Kazakov
Mixture-of-Experts (MoE) models repeatedly route tokens to sparse subsets of experts, but conventional routers expose no routing-specific record of how cross-layer influences accum…
Accepted Prefixes Are Not All You Need: A Negative Result on PEFT-Based Block-Diffusion Drafting
Abdurrahman Javat, Allan Kazakov
Speculative decoding accelerates autoregressive language model inference by using a cheap drafter to propose multiple future tokens and a target model to verify them. A common desi…
Revitalizing Dense Material Segmentation: Stabilized Vision Transformers and the Generalization Paradox
Allan Kazakov, Duygu Cakir, Hilal Kurt İrfanoğlu +1
Material segmentation, the pixel-wise classification of physical surface properties, remains a challenging problem in computer vision, requiring physicochemical understanding disti…
Silicon Showdown: Performance, Efficiency, and Ecosystem Barriers in Consumer-Grade LLM Inference
Abdurrahman Javat, Allan Kazakov
The operational landscape of local Large Language Model (LLM) inference has shifted from lightweight models to datacenter-class weights exceeding 70B parameters, creating profound…