Is TokenTiming superseded?

TokenTiming (Speculative decoding): current frontier — recent, not yet superseded in the knowledge base. 0 paper(s) critique it, 0 beat it on benchmarks — not ranked as a superseded baseline. Sub-problem: cluster led by EAGLE-2. Newer alternatives in the same sub-problem include DREAM-S, PPOW, SpecForge, OnlineSpec, MoE-Spec.

Speculative decoding

TokenTiming

TokenTiming: A Dynamic Alignment Method for Universal Speculative Decoding Model Pairs

On the frontierfirst seen Oct 17, 2025

Current frontier — recent, not yet superseded in the knowledge base

0 papers critique it · 0 beat it on benchmarks

What to use instead

Recent methods in the same sub-problem, not yet superseded in the knowledge base — arXiv benchmark leaders, not vetted production recommendations.

DREAM-SDREAM-S: Speculative Decoding with Searchable Drafting and Target-Aware Refinement for Multimodal Generation
May 30, 2026
PPOWPerformance-Driven Policy Optimization for Speculative Decoding with Adaptive Windowing
May 14, 2026
SpecForgeSpecForge: A Flexible and Efficient Open-Source Training Framework for Speculative Decoding
Mar 19, 2026
OnlineSpecWhen Drafts Evolve: Speculative Decoding Meets Online Learning
Mar 13, 2026
MoE-SpecMoE-Spec: Expert Budgeting for Efficient Speculative Decoding
Feb 17, 2026
AdaSPECAdaSPEC: Selective Knowledge Distillation for Efficient Speculative Decoders
Oct 22, 2025
Hierarchical Speculative Decoding (HSD)Fast Inference via Hierarchical Speculative Decoding
Oct 22, 2025
TokenTimingTokenTiming: A Dynamic Alignment Method for Universal Speculative Decoding Model Pairs
Oct 17, 2025
Draft, Verify, & Improve (DVI)Draft, Verify, and Improve: Toward Training-Aware Speculative Decoding
Oct 6, 2025
FastGRPOFastGRPO: Accelerating Policy Optimization via Concurrency-aware Speculative Decoding and Online Draft Learning
Sep 26, 2025