No article text was recorded for this version.
speculative decoding complexity
Version 1 of 171 papers74 conceptsfrozen 2026-10-04
Papers in this version
- Blog: SWIFT: On-the-Fly Self-Speculative Decoding for LLM Inference Acceleration
- Blog: Turning Trash into Treasure: Accelerating Inference of Large Language Models with Token Recycling
- Blog: FIRP: Faster LLM inference via future intermediate representation prediction
- Blog: SSSD: Simply-Scalable Speculative Decoding
- Blog: SuffixDecoding: Extreme Speculative Decoding for Emerging AI Applications
- Blog: When Linear Attention Meets Autoregressive Decoding: Towards More Effective and Efficient Linearized Large Language Models
- Blog: Train Once, and Decode As You Like
- Blog: Speculative Coupled Decoding for Training-Free Lossless Acceleration of Autoregressive Visual Generation
- Blog: Lossless Acceleration of Large Language Models with Hierarchical Drafting based on Temporal Locality in Speculative Decoding
- Blog: SWIFT: On-the-Fly Self-Speculative Decoding for LLM Inference Acceleration
- Blog: CAS-Spec: Cascade Adaptive Self-Speculative Decoding for On-the-Fly Lossless Inference Acceleration of LLMs
- Blog: Diffusion Models are Secretly Exchangeable: Parallelizing DDPMs via Autospeculation
- Blog: AdaDecode: Accelerating LLM Decoding with Adaptive Layer Parallelism
- Blog: Faster In-Context Learning for LLMs via N-Gram Trie Speculative Decoding
- Blog: Make Some Noise: Unlocking Language Model Parallel Inference Capability through Noisy Training
- Blog: TokenSwift: Lossless Acceleration of Ultra Long Sequence Generation
- Blog: Diffusion Models are Secretly Exchangeable: Parallelizing DDPMs via Auto Speculation
- Blog: REST: Retrieval-Based Speculative Decoding
- Blog: Sparse-to-Dense: A Free Lunch for Lossless Acceleration of Video Understanding in LLMs
- Blog: Break the Sequential Dependency of LLM Inference Using Lookahead Decoding
- Blog: Multi-Token Prediction via Self-Distillation
- Blog: LEAP: Zone-Aware MCTS for LLM Self-Speculative Decoding
- Blog: Make Some Noise: Unlocking Language Model Parallel Inference Capability through Noisy Training
- Blog: PARD: Accelerating LLM Inference with Low‑Cost PARallel Draft Model Adaptation
- Blog: Mirror Speculative Decoding: Breaking the Serial Barrier in LLM Inference
- Blog: Cacheback: Speculative Decoding With Nothing But Cache
- Blog: Accelerating Auto-regressive Text-to-Image Generation with Training-free Speculative Jacobi Decoding
- Blog: FractalLLM: Lossless Self-Speculative Decoding with Layer Embedded Self-Compression
- Blog: Well Begun is Half Done: Low-resource Preference Alignment by Weak-to-Strong Decoding
- Blog: RetroLLM: Empowering Large Language Models to Retrieve Fine-grained Evidence within Generation
- Blog: MC-SJD : Maximal Coupling Speculative Jacobi Decoding for Autoregressive Visual Generation Acceleration
- Blog: Medusa: Simple LLM Inference Acceleration Framework with Multiple Decoding Heads
- Blog: DistillSpec: Improving Speculative Decoding via Knowledge Distillation
- Blog: SSSD: Simply-Scalable Speculative Decoding
- Blog: Efficient Parallel Samplers for Recurrent-Depth Models and Their Connection to Diffusion Language Models
- Blog: Adaptive Draft-Verification for Efficient Large Language Model Decoding
- Blog: SDFP: Speculative Decoding with FIT-Pruned Models for Training-Free and Plug-and-Play LLM Acceleration
- Blog: Speculative Decoding for Multi-Sample Inference
- Blog: Whisper in Medusa's Ear: Multi-head Efficient Decoding for Transformer-based ASR
- Blog: Kangaroo: Lossless Self-Speculative Decoding for Accelerating LLMs via Double Early Exiting
- Blog: EAGLE-2: Faster Inference of Language Models with Dynamic Draft Trees
- Blog: Inference with Reference: Lossless Acceleration of Large Language Models
- Blog: Lookahead: An Inference Acceleration Framework for Large Language Model with Lossless Generation Accuracy
- Blog: Multi-Branch Self-Drafting for LLM Inference Acceleration
- Blog: Structuring The Future: Diffusion LLM Speculative Decoding via Calibrated Draft Graphs
- Blog: Free Draft-and-Verification: Toward Lossless Parallel Decoding for Diffusion Large Language Models
- Blog: AdaEDL: Early Draft Stopping for Speculative Decoding of Large Language Models via an Entropy-based Lower Bound on Token Acceptance Probability
- Blog: Faster Speculative Decoding via Effective Draft Decoder with Pruned Candidate Tree
- Blog: Accelerated Diffusion Models via Speculative Sampling
- Blog: Novel Parasitic Dual-Scale Modeling for Efficient and Accurate Multilingual Speech Translation
- Blog: The N-Grammys: Accelerating Autoregressive Inference with Learning-Free Batched Speculation
- Blog: REST: Retrieval-Based Speculative Decoding
- Blog: PLD+: Accelerating LLM Inference by Leveraging Language Model Artifacts
- Blog: AdaDecode: Accelerating LLM Decoding with Adaptive Layer Parallelism
- Blog: Characterizing self-speculative decoding approaches for accelerating LLMs
- Blog: Lookahead: An Inference Acceleration Framework for Large Language Model with Lossless Generation Accuracy
- Blog: Gumiho: A Hybrid Architecture to Prioritize Early Tokens in Speculative Decoding
- Blog: DynaMo: Accelerating Language Model Inference with Dynamic Multi-Token Sampling
- Blog: SPECTRA: Faster Large Language Model Inference with Optimized Internal and External Speculation
- Blog: Fast and Accurate Causal Parallel Decoding using Jacobi Forcing
Showing 60 of 71 papers.
History & citation
Current · Permanent link
- This direction split from Training-Free and Self-Speculative Decoding for LLM Acceleration (66% overlap)
Cite this direction
Latest version (a living page):
@misc{lacuna_speculative_decoding_complexity_1d4f420c,
title = {speculative decoding complexity},
author = {{Lacuna}},
howpublished = {Lacuna research direction},
year = {2026},
url = {https://lacuna.tiptreesystems.com/direction/speculative-decoding-complexity/rdr_1d4f420c585545bbb3a647bc5ed6a6d0},
note = {Latest version of a living research direction; cite a version for a fixed text. Accessed 2026-11-20},
}This exact version (version 1, frozen text):
@misc{lacuna_speculative_decoding_complexity_1d4f420c_v1,
title = {speculative decoding complexity},
author = {{Lacuna}},
howpublished = {Lacuna research direction},
year = {2026},
url = {https://lacuna.tiptreesystems.com/direction/speculative-decoding-complexity/rdr_1d4f420c585545bbb3a647bc5ed6a6d0/v/1},
note = {Version 1, frozen 2026-10-04. Accessed 2026-11-20},
}