Future Token Prediction – Causal Language Modelling with Per-Token Semantic State Vector for Multi-Token Predictionby…
Multi-Draft Speculative Sampling: Canonical Architectures and Theoretical Limitsby Ashish Khisti, M.Reza Ebrahimi, Hassan Dbouk, Arash…
AdaEDL: Early Draft Stopping for Speculative Decoding of Large Language Models via an Entropy-based Lower…
Faster Language Models with Better Multi-Token Prediction Using Tensor Decompositionby Artem Basharin, Andrei Chertkov, Ivan…
AMUSD: Asynchronous Multi-Device Speculative Decoding for LLM Accelerationby Bradley McDanelFirst submitted to arxiv on: 22…
TreeBoN: Enhancing Inference-Time Alignment with Speculative Tree-Search and Best-of-N Samplingby Jiahao Qiu, Yifu Lu, Yifan…
Random Token Fusion for Multi-View Medical Diagnosisby Jingyu Guo, Christos Matsoukas, Fredrik Strand, Kevin SmithFirst…
EPIC: Efficient Position-Independent Context Caching for Serving Large Language Modelsby Junhao Hu, Wenrui Huang, Haoyi…
Faster-GCG: Efficient Discrete Optimization Jailbreak Attacks against Aligned Large Language Modelsby Xiao Li, Zhuhong Li,…
Is Less More? Exploring Token Condensation as Training-free Test-time Adaptationby Zixin Wang, Dong Gong, Sen…