The authors present RAGthoven, a system for SemEval-2026 Task 1 that decomposes creative text generation into a multi-stage large language model pipeline grounded in computational humor theory. The final configuration augments the Planner with retrieval-augmented generation from a curated joke corpus and evaluates agentic variants against a non-agentic baseline.
- RAGthoven shares Rank 1 with the Gemini 2.5 Flash baseline in English, Spanish, and Chinese, with overlapping confidence intervals.
- In Spanish, it leads the baseline by 42 raw Elo points (1182 vs. 1140).
- In English (1045 vs. 1081) and Chinese (1045 vs. 1053), the baseline holds the higher raw rating within the same statistical tie.
- Agentic variants using ReAct-style tool-calling or autonomous orchestration did not produce superior outputs despite higher tool-call budgets.
The results suggest language-dependent diminishing returns from elaborate multi-stage prompt engineering and agentic scaffolding once a strong frontier model is in the loop.