AI Paper Atlas

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

Parallel Token Generation for Language Models

Felix Draxler, Justus Will, Farrin Marouf Sofian, Theofanis Karaletsos, Sameer Singh, Stephan Mandt

ICLR 2026 · Large language models › Language, safety and interpretability · approach: Parallel and early-exit decoding

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
condition draft tokens on auxiliary variables instead of independent prediction (ablation)O-PTP draft (TinyLlama-1.1B) · CodeContests speculative decodingaccepted tokens per step: 6.2 -> 7.0Improved
use parallel instead of sequential verification (ablation)O-PTP · SpecBenchaccepted tokens per step: 4.21 -> 2.89Hurt
distill O-PTP versus prior speculative methodsO-PTP (Vicuna-7B) · SpecBench task averagewall-clock speedup: 2.40 vs 2.34 (SAMD)Improved
train C-PTP from data alone versus autoregressive modelC-PTP · Taxi pickup sequencesperplexity: 19.88 vs 19.81No clear effect
use structured (ar + lin) auxiliary embedding instead of fully learned (fl + NN) (ablation)O-PTP · Taxiaverage correct tokens: 35.9 -> 45.4Improved
use O-PTP instead of C-PTP (ablation)PTP · Taxiaverage correct tokens: 40.4 -> 45.4Improved
increase number of parallel proposal sequences M (ablation)O-PTP · Taxi N=100average correct tokens: 45.36 -> 59.79Improved
limit tokens predicted per call N (ablation)O-PTP · Taxiaverage correct tokens: 45.36 (N=100) vs 13.90 (N=16)Hurt
vary Beta parameter b for sampling training auxiliaries (ablation)O-PTP · Taxi N=16average correct tokens: 12.9 / 13.9 / 13.8No clear effect
PTP draft instead of autoregressive draft at equal sizePTP draft · CodeContests speculative decodingspeedupImproved
sample training sequences from teacher vs dataset (ablation)O-PTP kl · Taxi N=16average correct tokens: 29 -> 40Improved

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