Polymathic AI and Princeton researchers release FlashMap, a flow map drafter for faster LLM decoding
AIResearchers from Polymathic AI, Princeton University, and other institutions released FlashMap, a speculative decoding drafter that proposes a 16-token block in one forward pass. The research team reports up to a 6.14x lossless speedup on GSM8K with Qwen3-8B, with drafter weights on Hugging Face and vLLM and SGLang plugins.
AIWhy it matters.
The article compares FlashMap with DFlash and EAGLE-3 on matched retrained baselines, showing where its speedup and acceptance gains come from.
Source details
MarkTechPost · Asif Razzaq · marktechpost.comPublished · added here
Original title: Polymathic AI and Princeton Researchers Release FlashMap, a Flow Map Drafter for 6.14x Faster LLM Decoding