Execution feedback in refinement loops improves 1-3B code generation performance far more than complex pipeline topologies discovered via evolutionary search on HumanEval and sanitized MBPP.
Ares: Adaptive red-teaming and end-to-end repair of policy-reward system
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.SE 1years
2026 1verdicts
ACCEPT 1representative citing papers
citing papers explorer
-
Feedback Over Form: Why Execution Feedback Matters More Than Pipeline Topology in 1-3B Code Generation
Execution feedback in refinement loops improves 1-3B code generation performance far more than complex pipeline topologies discovered via evolutionary search on HumanEval and sanitized MBPP.