AI Paper Atlas

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

EvoVLA: Self-Evolving Vision-Language-Action Model

Zeting Liu, ZIDA YANG, Zeyu Zhang, Hao Tang

ECCV 2026 · Vision-language and multimodal › Robot action and navigation · approach: Vision-language-action foundation models · 13 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 counterfactual hard negatives to SAR (ablation)OpenVLA-OFT · Discoverse-L avgsuccess rate (%): 59.0 -> 61.8Improved
add counterfactual hard negatives to SAR (ablation)OpenVLA-OFT · Discoverse-L avghallucination rate (%): 38.5 -> 31.2Improved
add temporal smoothing of stage scores (ablation)OpenVLA-OFT+hard negatives · Discoverse-L avgsuccess rate (%): 61.8 -> 63.7Improved
add temporal smoothing of stage scores (ablation)OpenVLA-OFT+hard negatives · Discoverse-L avghallucination rate (%): 31.2 -> 23.4Improved
add Long-Horizon Memory (ablation)EvoVLA partial · Discoverse-L avgsuccess rate (%): 63.7 -> 66.1Improved
add Long-Horizon Memory (ablation)EvoVLA partial · Discoverse-L avghallucination rate (%): 23.4 -> 19.5Improved
add pose-based object exploration (ablation)EvoVLA partial · Discoverse-L avgsuccess rate (%): 66.1 -> 69.2Improved
add pose-based object exploration (ablation)EvoVLA partial · Discoverse-L avghallucination rate (%): 19.5 -> 14.8Improved
EvoVLA modules vs OpenVLA-OFT backboneEvoVLA · Discoverse-L (sim)success rate (%): 59.0 -> 69.2Improved
EvoVLA vs OpenVLA-OFT on real robotEvoVLA · Real-world 4 tasks avgsuccess rate (%): 43.6 -> 54.6Improved
intrinsic weight rho 0.6 vs 0.3 (ablation)EvoVLA · Discoverse-L avgsuccess rate (%): 67.1 -> 69.2Improved
intrinsic weight rho 0.6 vs 0.9 (ablation)EvoVLA · Discoverse-L avgsuccess rate (%): 66.8 -> 69.2Improved

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