Search papers, labs, and topics across Lattice.
This paper addresses the challenge of maintaining software quality in AI-generated code by proposing a method for redundant-code reduction through proposal scheduling. The authors introduce DELSCOUT, which ranks deletion candidates based on a budgeted execution-verification capacity, demonstrating that the order of candidates significantly influences the effectiveness of code deletion. Their approach achieves a 9.5% increase in verified-deletion coverage while using fewer verifier calls compared to static baselines, highlighting the importance of candidate ordering in the code deletion process.
Candidate ordering, not just model confidence, is the key to effective code deletion in AI systems, leading to a significant boost in verified coverage.
Frontier coding models now match or exceed strong human reference points on programming benchmarks, yet benchmark success does not imply maintainable software. Prompt-driven"vibe coding"is additive: new branches, guards, and fallbacks accumulate faster than obsolete logic is removed. We study the inverse problem-how an Al system should remove code when execution-verification capacity is finite. We formulate redundant-code reduction as proposal scheduling: a ranker orders single-statement deletion candidates, an execution suite accepts the first candidate that passes, and a budget bounds how many candidates may be tested. Our central observation is that candidate order, not model confidence, is the control surface a deployment can reason about. DELSCOUT instantiates two schedules. Given representative target-domain validation, a five-slot budget spends three slots on deterministic shortest-first candidates and two on complementary learned candidates; across nine MBPP replications with 0.5B, 0.6B, and 8B rankers this raises verified-deletion coverage by 9.5% relative (+6.7 accepted tasks) while consuming slightly fewer verifier calls than the matched static baseline. Without such validation the same rankers can lose coverage under shift, so we instead evaluate the complete static prefix first and append learned candidates only afterwards; for a deterministic verifier this makes coverage and character reduction non-decreasing by construction, at a measured 4.8-62.5% increase in verifier calls. MBPP+ then erases the in-domain advantage, showing that scheduling governs search while the test suite alone governs what"preserving behavior"means. The result is an auditable division of labor: models widen the search for removable code, order bounds the damage a mis-ranked proposal can do, and execution retains authority over every committed deletion.
