Memory Networks: The Multi-hop Update
Beginner
The problem, from first principles
Memory Networks reason over several hops: each hop reads memory with the current state, then updates the state with the read result. Later hops can follow chains of facts.
From theory to code
Implement hop_update(u, o), the state update between hops.
Constraints
- The update is a simple sum.
Hints
Hint 1
Add the memory output to the current state.
Keywords
research-paperssequence-modelsmemoryquestion-answering
Python
Run your code to see results
No results yet
Run your code to check it against the examples, or submit to run every test.