Clasp In Context Layer Skip For Self Speculative Decoding Acl Anthology
CLaSp: In-Context Layer Skip for Self-Speculative Decoding - ACL Anthology
DReSD: Dense Retrieval for Speculative Decoding - ACL Anthology
Speculative Decoding for Multi-Sample Inference - ACL Anthology
Mamba Drafters for Speculative Decoding - ACL Anthology
Paper page - CLaSp: In-Context Layer Skip for Self-Speculative Decoding
Graph-Structured Speculative Decoding - ACL Anthology
SAM Decoding: Speculative Decoding via Suffix Automaton - ACL Anthology
Hierarchical Speculative Decoding with Dynamic Window - ACL Anthology
Speculative Contrastive Decoding - ACL Anthology
Decoding Speculative Decoding - ACL Anthology
Advertisement Space (300x250)
How Speculative Can Speculative Decoding Be? - ACL Anthology
Speculative Safety-Aware Decoding - ACL Anthology
SLiM: Speculative Decoding with Hypothesis Reduction - ACL Anthology
A Drop-In Solution for On-the-Fly Adaptation of Speculative Decoding in ...
Critic-Guided Decoding for Controlled Text Generation - ACL Anthology
RASD: Retrieval-Augmented Speculative Decoding - ACL Anthology
Cacheback: Speculative Decoding With Nothing But Cache - ACL Anthology
Rethinking Skip Connection with Layer Normalization - ACL Anthology
Look-back Decoding for Open-Ended Text Generation - ACL Anthology
[论文评述] ConfLayers: Adaptive Confidence-based Layer Skipping for Self ...
Advertisement Space (336x280)
FractalLLM: Lossless Self-Speculative Decoding with Layer Embedded Self ...
Spec-VLA: Speculative Decoding for Vision-Language-Action Models with ...
Logically Constrained Decoding - ACL Anthology
EMS-SD: Efficient Multi-sample Speculative Decoding for Accelerating ...
An Empirical Study of Speculative Decoding for Small Language Models ...
Scaling Context Space - ACL Anthology
Dovetail: A CPU/GPU Heterogeneous Speculative Decoding for LLM ...
OPT-Tree: Speculative Decoding with Adaptive Draft Tree Structure - ACL ...
UniSpec: Training-Free Speculative Decoding for Robust LLM Acceleration ...
DSVD: Dynamic Self-Verify Decoding for Faithful Generation in Large ...
Advertisement Space (336x280)
Speculative Decoding via Early-exiting for Faster LLM Inference with ...
Adaptive Contrastive Decoding in Retrieval-Augmented Generation for ...
Lazy-k Decoding: Constrained Decoding for Information Extraction - ACL ...
Speculative Decoding in vLLM | OpenLM.ai
Speculative Decoding 论文阅读合订本 - 知乎
KnapSpec: Self-Speculative Decoding via Adaptive Layer Selection as a ...