No article text was recorded for this version.
speculative decoding
Version 2 of 2642 papers1,013 conceptsfrozen 2026-10-04
Papers in this version
- Blog: Inference with Reference: Lossless Acceleration of Large Language Models
- Blog: Free Draft-and-Verification: Toward Lossless Parallel Decoding for Diffusion Large Language Models
- Blog: PipeInfer: Accelerating LLM Inference using Asynchronous Pipelined Speculation
- Blog: Speculative Ensemble: Fast Large Language Model Ensemble via Speculation
- Blog: Pipeline Parallelism is All You Need for Optimized Early-Exit Based Self-Speculative Decoding
- Blog: Distributed Speculative Inference (DSI): Speculation Parallelism for Provably Faster Lossless Language Model Inference
- Blog: Lossless Acceleration for Seq2seq Generation with Aggressive Decoding
- Blog: SpecInfer: Accelerating Large Language Model Serving with Tree-based Speculative Inference and Verification
- Blog: Lossless Acceleration of Large Language Model via Adaptive N-gram Parallel Decoding
- Blog: Hardware-Aware Parallel Prompt Decoding for Memory-Efficient Acceleration of LLM Inference
- Blog: Self-Speculative Decoding Accelerates Lossless Inference in Any-Order and Any-Subset Autoregressive Models
- Blog: FlowSpec: Continuous Pipelined Speculative Decoding for Efficient Distributed LLM Inference
- Blog: Tutorial Proposal: Speculative Decoding for Efficient LLM Inference
- Blog: Blockwise Parallel Decoding for Deep Autoregressive Models
- Blog: Fast Large Language Model Collaborative Decoding via Speculation
- Blog: Self Speculative Decoding for Diffusion Large Language Models
- Blog: FractalLLM: Lossless Self-Speculative Decoding with Layer Embedded Self-Compression
- Blog: A Survey on Parallel Text Generation: From Parallel Decoding to Diffusion Language Models
- Blog: Snakes and Ladders: Accelerating SSM Inference with Speculative Decoding
- Blog: Fail Fast, Win Big: Rethinking the Drafting Strategy in Speculative Decoding via Diffusion LLMs
- Blog: PaSS: Parallel Speculative Sampling
- Blog: Distributed Speculative Inference (DSI): Speculation Parallelism for Provably Faster Lossless Language Model Inference
- Blog: Novel Parasitic Dual-Scale Modeling for Efficient and Accurate Multilingual Speech Translation
- Blog: L-MTP: Leap Multi-Token Prediction Beyond Adjacent Context for Large Language Models
- Blog: Fast and High-Quality Auto-Regressive Speech Synthesis via Speculative Decoding
- Blog: DuoDecoding: Hardware-aware Heterogeneous Speculative Decoding with Dynamic Multi-Sequence Drafting
- Blog: Nearest Neighbor Speculative Decoding for LLM Generation and Attribution
- Blog: FAFO: Lossy KV Cache Compression for Lossless Inference Acceleration via Draftless Fumble Decoding
- Blog: DualDiffusion: A Speculative Decoding Strategy for Masked Diffusion Models
- Blog: Predictive Pipelined Decoding: A Compute-Latency Trade-off for Exact LLM Decoding
- Blog: SJD++: Improved Speculative Jacobi Decoding for Training-free Acceleration of Discrete Auto-regressive Text-to-Image Generation
- Blog: Ouroboros: Generating Longer Drafts Phrase by Phrase for Faster Speculative Decoding
- Blog: Ouroboros: Generating Longer Drafts Phrase by Phrase for Faster Speculative Decoding
- Blog: Fast and scalable retrosynthetic planning with a transformer neural network and speculative beam search
- Blog: Faster In-Context Learning for LLMs via N-Gram Trie Speculative Decoding
- Blog: Novel Parasitic Dual-Scale Modeling for Efficient and Accurate Multilingual Speech Translation
- Blog: Big Little Transformer Decoder
- Blog: Lossless Speedup of Autoregressive Translation with Generalized Aggressive Decoding
- Blog: Parallel Speculative Decoding with Adaptive Draft Length
- Blog: Fast Inference from Transformers via Speculative Decoding
- Blog: Break the Sequential Dependency of LLM Inference Using Lookahead Decoding
- Blog: Speeding up Speculative Decoding via Approximate Verification
- Blog: AdaServe: SLO-Customized LLM Serving with Fine-Grained Speculative Decoding
- Blog: PARD: Accelerating LLM Inference with Low‑Cost PARallel Draft Model Adaptation
- Blog: PEARL: Parallel Speculative Decoding with Adaptive Draft Length
- Blog: Reviving Any-Subset Autoregressive Models with Principled Parallel Sampling and Speculative Decoding
- Blog: Interactive Text Games: Lookahead Is All You Need!
- Blog: Fast Inference via Hierarchical Speculative Decoding
- Blog: Whisper in Medusa's Ear: Multi-head Efficient Decoding for Transformer-based ASR
- Blog: Speculative Decoding: Exploiting Speculative Execution for Accelerating Seq2seq Generation
- Blog: SPECS: Faster Test-Time Scaling through Speculative Drafts
- Blog: AI Flow at the Network Edge
- Blog: Multi-Branch Self-Drafting for LLM Inference Acceleration
- Blog: QSpec: Speculative Decoding with Complementary Quantization Schemes
- Blog: A Comprehensive Survey of Accelerated Generation Techniques in Large Language Models
- Blog: Speculative Decoding via Early-exiting for Faster LLM Inference with Thompson Sampling Control Mechanism
- Blog: Predictive Sampling with Forecasting Autoregressive Models
- Blog: The N-Grammys: Accelerating Autoregressive Inference with Learning-Free Batched Speculation
- Blog: FLASH: Latent-Aware Semi-Autoregressive Speculative Decoding for Multimodal Tasks
- Blog: Sparse-to-Dense: A Free Lunch for Lossless Acceleration of Video Understanding in LLMs
Showing 60 of 642 papers.
History & citation
Current · Permanent link
- Lossless Parallel Inference via Speculative Decoding merged into this direction (75% overlap)
- Asynchronous Speculative Decoding for Parallel LLM Inference merged into this direction (56% overlap)
- Hybrid Speculative Decoding for Efficient Large Language Models merged into this direction (60% overlap)
- Multi-Token Prediction for Accelerated Autoregressive Inference merged into this direction (59% overlap)
- Training-Free and Self-Speculative Decoding for LLM Acceleration merged into this direction (51% overlap)
- Speculative Decoding for Efficient LLM Inference merged into this direction (64% overlap)
Cite this direction
Latest version (a living page):
@misc{lacuna_speculative_decoding_c8a35a73,
title = {speculative decoding},
author = {{Lacuna}},
howpublished = {Lacuna research direction},
year = {2026},
url = {https://lacuna.tiptreesystems.com/direction/speculative-decoding/rdr_c8a35a737b0d4ff68d7894c3848cbea4},
note = {Latest version of a living research direction; cite a version for a fixed text. Accessed 2026-11-20},
}This exact version (version 2, frozen text):
@misc{lacuna_speculative_decoding_c8a35a73_v2,
title = {speculative decoding},
author = {{Lacuna}},
howpublished = {Lacuna research direction},
year = {2026},
url = {https://lacuna.tiptreesystems.com/direction/speculative-decoding/rdr_c8a35a737b0d4ff68d7894c3848cbea4/v/2},
note = {Version 2, frozen 2026-10-04. Accessed 2026-11-20},
}