ResearchBenchmarksRebuttal Generation on RESEARCHARCADEFollow70.3SBERT SimilarityLLM60.00462.67765.3568.023Nov 27, 2025Evaluation ResultsMethodMethodLinksSBERT SimilarityRouge-LGPT-4o-mini ScoreLLMBackbone=GPTOSS-120B,...Backbone=GPTOSS-120B, Training=Fixed, Hop=1-hop2025.1170.315.288.4LLMBackbone=Qwen3-8B, Tra...Backbone=Qwen3-8B, Training=Fixed, Hop=1-hop2025.117015.420.8LLMBackbone=Qwen3-0.6B, T...Backbone=Qwen3-0.6B, Training=Trained, Hop=1-hop2025.1163.813.12.2LLMBackbone=Qwen3-0.6B, T...Backbone=Qwen3-0.6B, Training=Fixed, Hop=1-hop2025.1160.412.51.1