Simple neural policy variants can match programmatic policies on several OOD generalization benchmarks, challenging the claim that programmatic representations are inherently better at generalizing.
Qwen technical report, 2023
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Common Benchmarks Undervalue the Generalization Power of Programmatic Policies
Simple neural policy variants can match programmatic policies on several OOD generalization benchmarks, challenging the claim that programmatic representations are inherently better at generalizing.