AI Paper Atlas

Research map / Large language models / Language, safety and interpretability

Fast and Expressive Multi-Byte Prediction with Probabilistic Circuits

Andreas Grivas, Lorenzo Loconte, Emile van Krieken, Piotr Nawrot, Yu Zhao, Euan Wielewski, Pasquale Minervini, Edoardo Ponti, Antonio Vergari

ICML 2026 · Large language models › Language, safety and interpretability · approach: Parallel and early-exit decoding · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
increase CP mixture components r from 1 (FF) to 128 (ablation)MTPC-CP · EvaByte, speculative sampling, n=8mean accepted tokens: 5.19 -> 6.02Improved
increase CP mixture components r from 32 to 128 (ablation)MTPC-CP · EvaByte, speculative sampling, n=8throughput (tok/s): 191.0 -> 182.3Hurt
use CP mixture (r=32) instead of fully factorised heads (ablation)EvaByte-MTP · EvaByte, speculative sampling, n=8throughput (tok/s): 176.3 -> 191.0Improved
use HMM circuit head instead of BTree (ablation)MTPC · EvaByte and Llama, n=8 and n=16throughputHurt
extend MTP window from n=8 to n=16 with BTree head (ablation)MTPC-BTree · EvaByte and Llama, speculative samplingthroughputMixed
use expressive circuits instead of factorised head (ablation)EvaByte-MTP · EvaByte, greedy speculative decodingthroughputHurt
add LoRA layers to the draft backbone (n=16) (ablation)MTPC-BTree · EvaByte and Llama, speculative sampling and greedymean accepted tokens: +.69 with 4 LoRA layers (EvaByte, sampling)Improved
use 1 LoRA layer rather than 4 for EvaByte (ablation)MTPC-BTree · EvaByte, n=16throughputImproved
add LoRA layers to EvaByte-MTP with n=8 (ablation)EvaByte-MTP (FF) · EvaByte, n=8mean accepted tokensNo clear effect
train LoRA adapters on last 16 of 32 layers of EvaByte (ablation)EvaByte draft · EvaBytemean accepted tokens vs latency: +37% accepted, latency 1.5x forward passMixed
train more than 16 layers with LoRA (ablation)EvaByte draft · EvaByteacceptance rateNo clear effect
replace fully factorised MTP with BTree PC head plus LoRA, within speculative decodingEvaByte-MTP · Tulu3 validation, speculative samplingspeedup: 1.17x over best FF MTPImproved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy