CORTEXA
← Browse
arxivcs.LGcs.AI2026-07-09

A Practical Investigation of Training-free Relaxed Speculative Decoding

Guoxuan Xia, Luka Ribar, Paul Balanca

Speculative decoding accelerates sampling from an autoregressive LLM by using a faster auxiliary model to draft tokens which are then verified in parallel by the LLM. Standard speculative decoding is lossless: its rejection and resampling steps exactly preserve the LLM's sampling distribution. Recent work argues that relaxing this strict guarantee can yield further speed-ups, controlled capability-speed trade-offs, or even capability gains. We practically investigate training-free relaxed speculative decoding techniques, unify existing approaches within a shared framework, benchmark them on contemporary settings, and distil takeaways and empirical findings for practitioners. Important takeaways include: relaxation can require considerable capability evaluation unlike lossless speculative decoding, and many relaxed approaches rely on a drafter that is a good language model, making them unsuited for lightweight dedicated multi-token-prediction drafters.

View free PDFSource page

Related papers

arxivcs.CVcs.AIcs.LG2026-07-03

SPLIT: Training-Free AI-Generated and Partially Edited Video Detection via Spatial Patch-Level Incoherence and Temporal Roughness

Jongyeop Hyun, Hyounghun Kim

Deploying AI-generated video detectors in real-world services demands an ultra-low false positive rate (FPR) on real videos to avoid falsely rejecting authentic content, a regime where standard metrics such as AUROC fail to reflect actual operating behavior. We introduce Spatial…

View free PDFSource page
arxivcs.LGcs.AI2026-07-20

CoCurve: Cross-Module Co-Pruning Curvature for Training-Free Structured LLM Pruning

Zhiren Gong, Zihao Zeng, Zijie Wang, Tiantong Wang, Chau Yuen, Wei Yang Bryan Lim

Structured pruning compresses large language models (LLMs) by removing whole computational units, such as attention heads and feed-forward (FFN) channel groups. Most training-free methods, however, rank these units independently, implicitly treating the loss from pruning a set as…

View free PDFSource page
arxivcs.AIcs.CLcs.CYcs.LG2026-07-04

Explainable AI for Screening Abuse-Related Trauma in Bangladeshi Children: A Training-Free Multimodal Framework Evaluated on Noise-Aware Synthetic Data

Salma Hoque Talukdar Koli, Fahima Haque Talukder Jely

Bangladesh has an estimated 1.17 mental-health professionals per 100,000 population and only six child psychiatrists nationwide. No Bengali-language, culturally adapted tool exists for early screening of abuse-related psychological trauma in children. We present ShishuRaksha AI,…

View free PDFSource page
arxivcs.LGcs.AIcs.CR2026-07-02

kNNGuard: Turning LLM Hidden Activations into a Training-Free Configurable Guardrail

Mahmoud Abdelfattah, Hamid Nasiri, Peter Garraghan

Large language models (LLMs) are increasingly deployed in domains requiring guardrails to detect unsafe, off-topic, or adversarial prompts. Existing guardrails predominantly rely on fine-tuning to build classifiers, which often suffer from low generalization and high inference la…

View free PDFSource page