AI Paper Atlas

Research map / Large language models / Reasoning and chain of thought

SMART: When is it Actually Worth Expanding a Speculative Tree?

Lifu Wang, Pan ZHOU

ECCV 2026 · Large language models › Reasoning and chain of thought · approach: Speculative decoding · 3 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
add SMART marginal benefit-cost tree expansion to MSDMSD, MLLMs · MLLM benchmarks T=0avg speedup: +20.2% avg across three MLLMs (LLaVA-1.5 7B 1.18 -> 1.53)Improved
add SMART to MSDMSD, MLLMs · MLLM benchmarks T=1avg speedup: +19.9% avg; LLaVA-1.5 7B 1.77 -> 2.28Improved
add SMART to EAGLE-3EAGLE-3, LLMs · MT-bench/HumanEval/GSM8K T=0avg speedup: +19.6% avg; LLaMA-3.3 70B 2.69 -> 3.35Improved
add SMART to EAGLE-3EAGLE-3, LLMs · MT-bench/HumanEval/GSM8K T=1avg speedup: +15.4% avg; LLaMA-3.1 8B 2.11 -> 2.38Improved
add SMART to MSD at batch size 1 (ablation)MSD · LLaVA-1.5 RTX Pro 6000 batch 1speedup: 2.20 -> 2.17No clear effect
add SMART to MSD at batch size 16 (ablation)MSD · LLaVA-1.5 RTX Pro 6000 batch 16speedup: 1.22 -> 1.58Improved
add SMART to MSD at batch size 32 (ablation)MSD · LLaVA-1.5 RTX Pro 6000 batch 32speedup: 0.82 -> 1.39Improved
add SMART to MSD at batch size 12 (ablation)MSD · LLaVA-1.5 L40S batch 12speedup: 0.90 -> 1.40Improved
increase verification token budget from 200 to 300-400 (ablation)SMART · RTX Pro 6000 batch 16, T=0avg speedup: 1.58 -> 1.28 (300), 1.27 (400)Hurt
decrease token budget from 200 to 100 (ablation)SMART · RTX Pro 6000 batch 16, T=0avg speedup: 1.58 -> 1.43Hurt
set discount factor alpha=1.0 instead of 0.8 (ablation)SMART · RTX Pro 6000 batch 16, T=0avg speedup: 1.56 -> 1.51Hurt
vary discount factor alpha within 0.5-0.9 (ablation)SMART · RTX Pro 6000 batch 16, T=0avg speedup: 1.54 to 1.56No clear effect

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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