paper_type.json (267B)
1 { 2 "paper_type": "empirical", 3 "reason": "Runs experiments merging task-specific LoRA adapters and reports quantitative results (pass@1/pass@10 improvements) on HumanEvalFix, with primary contribution being experimental findings about adapter merging strategies." 4 }