AI Paper Atlas

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

SCALE: Self-uncertainty Conditioned Adaptive Looking and Execution for Vision-Language-Action Models

Hyeonbeom Choi, Daechul Ahn, Youhan Lee, Taewook Kang, Seongwon Cho, Jonghyun Choi

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 adaptive action decoding (ablation)SCALE · LIBERO-Long (OpenVLA)success rate (%): 52.7 -> 58.0Improved
add adaptive visual attention (ablation)SCALE · LIBERO-Long (OpenVLA)success rate (%): 52.7 -> 56.0Improved
combine adaptive decoding and adaptive visual attention (ablation)SCALE · LIBERO-Long (OpenVLA)success rate (%): 52.7 -> 63.3Improved
use dual-reference self-uncertainty instead of Self-Certainty (ablation)SCALE · LIBERO-Long (OpenVLA)success rate (%): 53.8 -> 63.3Improved
use self-uncertainty instead of entropy (ablation)SCALE · LIBERO-Long (OpenVLA)success rate (%): 55.4 -> 63.3Improved
modulate vision encoder attention instead of VLA cross-modal attention (ablation)SCALE · LIBERO-Long (OpenVLA)success rate (%): 57.4 -> 63.3Improved
use continuous adaptive scaling instead of binary fixed switch (ablation)SCALE visual modulation · LIBERO-Long (OpenVLA)success rate (%): 54.8 -> 57.4Improved
use uncertainty change from EMA instead of instantaneous uncertainty (ablation)SCALE visual modulation · LIBERO-Long (OpenVLA)success rate (%): 55.4 -> 63.3Improved
apply fixed-hyperparameter sampling (temperature, top-k, top-p) instead of greedypi0-FAST · LIBERO (Avg.)success rate (%): 91.2 -> 84.3/88.1/86.2Hurt
apply fixed-hyperparameter sampling instead of greedyOpenVLA · LIBERO (Avg.)success rate (%): 75.7 -> 76.7/76.7/77.2Improved
add SCALE to OpenVLAOpenVLA · LIBERO (Avg.)success rate (%): 75.7 -> 81.5Improved
add SCALE to SpatialVLASpatialVLA (zero-shot) · SIMPLER-WidowX (Avg.)success rate (%): 31.3 -> 41.7Improved

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