The authors decomposed each long-horizon instruction into subtasks that could be grounded to verifiable simulator predicates, such as whether an object had been placed or a fixture operated.
Structured intermediate rewards improve robots on long, dependent tasks
StructRL rewards verified subtasks in prerequisite order, giving robot policies useful feedback before an entire manipulation sequence succeeds.
Big Tech
Ziyi Yin · Sangmin Woo · Kang Zhou · Sungyeon Kim · Aosong Feng · Haibo Ding · +1 more
The Pennsylvania State University · Amazon AWS AI
Research Digest··2 min read
Yin et al.
Why this paper
From Amazon AWS AI and The Pennsylvania State University · Released code
In one line
Structured intermediate rewards, gated by task prerequisites and scaled by completion pace, improve long-horizon VLA training over sparse terminal rewards.
What it released
Code
What we could check
- ✓Code link in the paper (github.com)
- ·No weights link found
- ·No dataset link found
- ·No compute details found
- ·No stated limitations found
- ✓Reports numbers on named benchmarks
Observed from the paper text and links we have. Absence here means we did not find it, not that it does not exist.
§