This week’s news from Zed, Anthropic, and OpenRouter shows why better harnesses matter more than better models
thenewstack.io·13h ago
AI Research
arXiv CS.AI·Jun 12·5 min read
EWAM: An Enhanced World Action Model for Closed-Loop Online Adaptation in Embodied Intelligence
TL;DR
Embodied intelligence systems often struggle with adapting to new tasks without extensive additional data. The Enhanced World Action Model (EWAM) was developed to enable closed-loop online adaptation using a pretrained Cosmos3 backbone without fine-tuning.
✦ Why It Matters
Engineers can leverage EWAM to create adaptable AI systems that require less data for new tasks.
Key Takeaways
- Utilize EWAM for real-time adaptation in robotic applications.
- Leverage pretrained models to avoid extensive retraining.
- Implement anomaly detection for improved task execution.
- Integrate lightweight neural layers for efficient performance.
How It Works
EWAM employs a co-reasoning mechanism that integrates four neural layers into the existing Cosmos3 architecture. The Neural Experience Memory Layer enhances task context, while the Neural Anomaly Detection Layer identifies discrepancies in real-time.
The Neural Policy Routing Layer dynamically selects the best response based on anomaly severity, and the Neural Action Correction Layer fine-tunes actions based on execution diagnostics. This architecture allows for seamless adaptation to new tasks without requiring additional data.