AI Paper Atlas

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

Uncovering Robot Vulnerabilities through Semantic Potential Fields

Som Sagar, Jiafei Duan, Sreevishakh Vasudevan, Yifan Zhou, Heni Ben Amor, Dieter Fox, Ransalu Senanayake

ICLR 2026 · Vision-language and multimodal › Robot action and navigation · approach: Memory and subgoal-based manipulation

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
use PPO instead of A2C or SAC as the vulnerability search policy (ablation)RoboMD · Failure ranking, 3 tasksavg accuracy score: 75.0 (A2C) / 52.2 (SAC) -> 80.7Improved
use RoboMD instead of VLMs for failure rankingRoboMD · Failure rankingaccuracy: VLMs below 60% vs PPO 76.0-84.0%Improved
add 5-shot in-context examples to GPT-4o (ablation)GPT-4o · Failure ranking, Squareaccuracy: 44.0% -> 48.6%Improved
add contrastive loss with image+text embedding (ablation)embedding · Embedding qualityMSE to identity: 0.8426 -> 0.1801Improved
add contrastive loss to image-only embedding (ablation)embedding · Embedding qualityMSE to identity: 0.6495 -> 0.6179No clear effect
add text (CLIP) encoder to image embedding with BCE only (ablation)embedding · Embedding qualityMSE to identity: 0.6495 -> 0.8426Hurt
fine-tune policy on RoboMD-selected failuresBC-lift policy · BC-liftaccuracy (%): 67.91 -> 92.83Improved
select failures with RoboMD instead of fine-tuning on all failures (ablation)BC-lift policy · BC-liftaccuracy (%): 85.48 -> 92.83 with 1.3 GB vs 9.0 GBImproved
compare model-based embeddings under lowering light intensityembedding · Lighting sweepcumulative embedding distance: 0.00 -> 0.67 monotonicImproved

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