Mistletoe introduces a stealthy attack on speculative decoding that collapses acceleration by reducing average accepted length while preserving output semantics.
hub Canonical reference
EAGLE: Speculative Sampling Requires Rethinking Feature Uncertainty
Canonical reference. 80% of citing Pith papers cite this work as background.
34
Pith papers citing it
Background
80% of classified citations
abstract
Autoregressive decoding makes the inference of Large Language Models (LLMs) time-consuming. In this paper, we reconsider speculative sampling and derive two key observations. Firstly, autoregression at the feature (second-to-top-layer) level is more straightforward than at the token level. Secondly, the inherent uncertainty in feature (second-to-top-layer) level autoregression constrains its performance. Based on these insights, we introduce EAGLE (Extrapolation Algorithm for Greater Language-model Efficiency), a simple yet highly efficient speculative sampling framework. By incorporating a token sequence advanced by one time step, EAGLE effectively resolves the uncertainty, enabling precise second-to-top-layer feature prediction with minimal overhead. We conducted comprehensive evaluations of EAGLE, including all models from the Vicuna and LLaMA2-Chat series, the MoE model Mixtral 8x7B Instruct, and tasks in dialogue, code generation, mathematical reasoning, and instruction following. For LLaMA2-Chat 70B, EAGLE achieved a latency speedup ratio of 2.7x-3.5x, doubled throughput, while maintaining the distribution of the generated text.
hub tools
citation-role summary
background 8
method 2
citation-polarity summary
representative citing papers
Graft combines pruning and retrieval in a sequential mechanism to build hybrid draft trees for speculative decoding, delivering up to 5.41× speedup and 21.8% better average speedup than EAGLE-3 on large models.
A new speculative inference system speeds up diffusion VLAs to 19.1 ms average latency (3.04x faster) on LIBERO by replacing most full 58 ms inferences with 7.8 ms draft rounds while preserving task performance.
SlimSpec replaces the standard LM-head in draft models with a low-rank version to deliver 4-5x faster speculative decoding while preserving full vocabulary and competitive acceptance rates.
BubbleSpec exploits long-tail bubbles in synchronous RL by using faster ranks' idle time to pre-generate rollout drafts for speculative decoding, reducing steps by 50% and raising throughput up to 1.8x while preserving exact synchrony.
NI Sampling accelerates discrete diffusion language models up to 14.3 times by training a neural indicator to select which tokens to sample at each step using a trajectory-preserving objective.
WISV uses a channel-aware semantic acceptance policy on hidden representations to boost accepted sequence length by up to 60.8% and cut interaction rounds by 37.3% in distributed speculative decoding, with under 1% accuracy loss.
Visual token pruning in MLLMs fails on complex reasoning due to Relevant Visual Information Shift during decoding, but the DSTP framework fixes it training-free across models.
DMax uses On-Policy Uniform Training and Soft Parallel Decoding to enable aggressive parallelism in dLLMs, raising TPF on GSM8K from 2.04 to 5.47 and on MBPP from 2.71 to 5.86 while preserving accuracy.
Drift-AR achieves 3.8-5.5x speedup in AR-diffusion image models by using entropy to enable entropy-informed speculative decoding and single-step (1-NFE) anti-symmetric drifting decoding.
KERV integrates kinematic Kalman Filter predictions with speculative decoding in VLA models to achieve 27-37% faster inference while maintaining nearly the same task success rates.
A Markov category framework for language models provides an information-theoretic rationale for speculative decoding and shows that a quadratic surrogate to negative log-likelihood induces generalized CCA alignment in linear-softmax heads after normalization.
VeriCache turns lossy KV cache compression into lossless LLM inference by drafting with compressed cache and verifying drafts with full cache, achieving up to 4x throughput with identical outputs.
PPOW uses window-level RL with cost-aware speedup and proximity rewards plus adaptive divergence-aware windowing to reach 6.29-6.52 acceptance lengths and 3.39-4.36x speedups in speculative decoding.
PARD-2 uses Confidence-Adaptive Token optimization to align draft model training with acceptance length in speculative decoding, enabling dual-mode operation and up to 6.94x lossless speedup on Llama3.1-8B.
CASCADE formalizes semantic interchangeability and convergence in target model representations to enable context-aware acceptance relaxation in tree-based speculative decoding, delivering up to 3.6x speedup on text-to-image models without quality loss.
CoVSpec achieves up to 2.21x higher throughput and over 96% lower communication overhead for device-edge VLM inference via training-free visual token reduction, adaptive drafting, and decoupled parallel verification-correction in speculative decoding.
EVICT adaptively truncates draft trees in MoE speculative decoding by combining drafter signals with profiled costs to retain only cost-effective prefixes, delivering up to 2.35x speedup over autoregressive decoding.
SPIN co-designs sparse attention with hierarchical memory to achieve 1.66-5.66x higher throughput, 7-9x lower TTFT, and up to 58% lower TPOT than vLLM and original sparse implementations.
NVLLM offloads FFN computations to integrated 3D NAND flash with page-level access and keeps attention in DRAM, delivering 16.7x-37.9x speedups over GPU out-of-core baselines for models up to 30B parameters.
SpecBound achieves up to 2.33x wall-time speedup in LLM inference via adaptive bounded self-speculation and layer-wise confidence calibration while preserving exact output equivalence.
SMART uses marginal benefit-cost analysis to dynamically build efficient speculative trees, achieving 15-20% additional speedup in LLM and MLLM inference.
SV-VLA uses infrequent heavy VLA planning of action chunks plus a lightweight closed-loop verifier to achieve both efficiency and robustness in dynamic robot control.
GlimpRouter uses the entropy of the first token in each reasoning step to decide whether to invoke a large model, yielding 10.7% higher accuracy and 25.9% lower latency than a standalone large model on AIME25.
citing papers explorer
No citing papers match the current filters.