Search papers, labs, and topics across Lattice.
RAGthoven is a multi-stage pipeline designed for multilingual humor generation, incorporating a Planner, a Best-of-N Writer, a Reflector for self-critique, and a Judge, all grounded in computational humor theory. Despite exploring two agentic variants that utilized higher tool-call budgets, the results showed no significant improvement over the non-agentic pipeline, with RAGthoven achieving competitive performance against the Gemini 2.5 Flash baseline across three languages. Notably, while it outperformed the baseline in Spanish, it fell short in English and Chinese, indicating that complex multi-stage architectures may yield diminishing returns in certain contexts.
Language-dependent diminishing returns challenge the efficacy of elaborate multi-stage humor generation pipelines in AI.
We present RAGthoven, our system for SemEval-2026 Task 1 (MWAHAHA), Subtask A (multilingual constrained humor generation in English, Spanish, and Chinese). RAGthoven decomposes creative text generation into a multi-stage large language model (LLM) pipeline (Planner, Best-of-N Writer, Reflector for self-critique, LLM-as-a-judge Judge) grounded in computational humor theory (Benign Violation Theory, Script-based Semantic Theory of Humor) and refined across ten experiments. In our final configuration, we augment the Planner with retrieval-augmented generation (RAG) from a curated joke corpus, seeding generation with diverse joke mechanisms. We also evaluate two agentic variants -- ReAct-style sequential tool-calling (Exp09) and autonomous multi-branch orchestration (Exp10) -- that expose the same four stages with a deterministic ConstraintAudit checker. Across four frontier models on a held-out 12-instance English sample, neither agentic variant produced outputs we judged superior to the non-agentic pipeline despite substantially higher tool-call budgets. RAGthoven shares Rank 1 with the Gemini 2.5 Flash baseline in all three languages, with overlapping organizer-reported confidence intervals. In Spanish, it leads the baseline by 42 raw Elo points (1182 vs. 1140), while in English (1045 vs. 1081) and Chinese (1045 vs. 1053) the baseline holds the higher raw rating within the same statistical tie. Together, these results suggest language-dependent diminishing returns from elaborate multi-stage prompt engineering and agentic scaffolding once a strong frontier model is in the loop.
