AI Paper Atlas

Research map / Vision-language and multimodal / Robot action and navigation

Think Less, Act Early: Reinforced Latent Reasoning with Early Exit in Vision-Language-Action Models

Dianqiao Lei, Lianlei Shan

ICML 2026 · Vision-language and multimodal › Robot action and navigation · approach: Vision-language-action foundation models · 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
add latent reasoning (ablation)AVA-VLA · LIBEROAverage success rate (%): 95.8 -> 98.3Improved
add RL denoising (ablation)AVA-VLA · LIBEROAverage success rate (%): 96.6 -> 98.3Improved
add latent smoothness regularization (ablation)AVA-VLA · LIBEROAverage success rate (%): 96.9 -> 98.3Improved
add early exit (ablation)AVA-VLA · LIBEROAverage success rate (%): 98.0 -> 98.3No clear effect
add early exit (ablation)AVA-VLA · CALVIN ABC->DAverage length: 4.61 -> 4.65Improved
add early exit (ablation)AVA-VLA · LIBERO-SpatialMean latency (ms): 312 -> 145Improved
add POMDP formulation and RL denoising to a BC latent module (ablation)AVA-VLA · LIBEROAverage success rate (%): 96.6 -> 98.3Improved
replace truncated explicit CoT with parallel latent reasoning at matched latency (ablation)AVA-VLA · LIBEROSuccess rate (%): 78.8 -> 98.3Improved
raise fixed reasoning depth from 5 to 10 (ablation)AVA-VLA · LIBEROSuccess rate (%): 98.0 -> 98.1No clear effect
use adaptive depth instead of direct single step (ablation)AVA-VLA · LIBEROSuccess rate (%): 94.2 -> 98.3Improved
lower early-exit threshold from 0.55 to 0.30 (ablation)AVA-VLA · LIBEROAverage success rate (%): 98.3 -> 96.1Hurt
add RL denoising (ablation)AVA-VLA · LIBERO+ (7 perturbations)Average success rate (%): 83.2 -> 89.3Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 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