Tags
paper-reading
2 articles
Harness VLA: Steering Frozen VLAs into Reliable Manipulation Primitives via Memory-Guided Agents
A paper note on using memory-guided LLM agents to turn frozen VLAs into retryable manipulation primitives.
Behavior-Skill: A Fine-Grained Benchmark for Evaluating Vision-Language-Action Policies in Long-Horizon Tasks
Notes on a benchmark that turns long-horizon VLA failures into skill-level measurements.