SPEAR: Selective Preference Elicitation with Adaptive Rewards for VLM-Guided Robotic Manipulation
Recent Vision-Language Models (VLMs) enable automated preference labeling for robotic manipulation, yet their deployment remains impractical since each query incurs substantial inference cost, while VLMs frequently abstain on visually challenging comparisons.
ProofPaper ↗
Key points
- We propose SPEAR, a framework that transforms VLMs into efficient preference oracles via two synergistic mechanisms.
- Second, we deploy an adaptive reward transition that bootstraps exploration with VLM-synthesized guidance before preference-based refinement.
- SPEAR learns three Mobile ALOHA manipulation tasks fully online on real hardware, demonstrating end-to-end real-world applicability.
- By addressing query efficiency and exploration challenges, SPEAR establishes VLMs as practical preference oracles for scalable robot learning without manual reward engineering.
Sources (1)
- [1]SPEAR: Selective Preference Elicitation with Adaptive Rewards for VLM-Guided Robotic ManipulationarXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 15, 10:02 AM
Recent Vision-Language Models (VLMs) enable automated preference labeling for robotic manipulation, yet their deployment remains impractical since each query incurs substantial inference cost, while VLMs frequently abstain on visually challenging comparisons.
We propose SPEAR, a framework that transforms VLMs into efficient preference oracles via two synergistic mechanisms.
Extractive summary: sentences quoted from the sources.
Before this
- Oct 15, 2026CQDC-Warp: GPU-Accelerated Differentiable Quasi-Dynamic Simulation for Contact-Rich Manipulation
- Oct 15, 2026Anthropic tells investors quarterly revenue up nearly 50% to $16.8bn ahead of IPO
- Oct 15, 2026Learning to Think and Explore in 3D World
- Oct 15, 2026Has GPT-6-Astra Achieved Embodied Intelligence?
- Oct 14, 2026Open-world AI: a Tale of Three Requirements
- Oct 14, 2026mistralai/Timestral-209M