Ablation study of correction methods in Polish language Retrieval-Augmented Generation chatbots
Computer Science and Mathematical Modelling · 15.11.2026 · 0 (23/2026) · 5-18
DOI: 10.5604/01.3001.0055.9903 ↗Kurz gefasst
Fachartikel in der Zeitschrift Computer Science and Mathematical Modelling, erschienen am 15.11.2026. Verfasst von Szymon Florek und 1 weiteren Autor:innen. Die Arbeit nennt 26 Literaturangaben.
Abstract (Originalsprache)
Retrieval-Augmented Generation (RAG) mitigates hallucinations in large language models by grounding responses in retrieved evidence, yet the interplay of correction techniques across the RAG pipeline remains underexplored – especially for non-English languages. We present a modular RAG system evaluated through
a systematic ablation study of 12 pipeline configurations on a 1,000-question Polish-language test set derived from a 1M-article Wikipedia corpus and manually validated. Four correction families are compared: (i) cross-
-encoder reranking, (ii) query decomposition with multi-hop reranking and diversity enforcement, (iii) native Chain-of-Thought (CoT) reasoning, and (iv) post-generative Chain-of-Verification (CoVe). Results evaluated with RAGAS metrics show that multi-hop reranking yields the largest quality gains (Δ Faithfulness +15.6%, Δ Context Recall +38.4%, Δ Relevancy +20.3%) at acceptable latency (18.4 s), while CoVe in auto-correction mode can paradoxically degrade faithfulness at high computational cost (+41.6 s). Cross-encoder reranking achieves the best cost quality ratio (+16.7% Recall, +1.2 s). We discuss practical configuration tiers for real-
-time, interactive, and offline deployments.
Zitieren
APA
Florek, S., & Górny, P. (2026). Ablation study of correction methods in Polish language Retrieval-Augmented Generation chatbots. Computer Science and Mathematical Modelling, 0(23/2026), 5-18. https://doi.org/10.5604/01.3001.0055.9903
Vancouver
Florek S, Górny P. Ablation study of correction methods in Polish language Retrieval-Augmented Generation chatbots. Computer Science and Mathematical Modelling. 2026;0(23/2026):5-18. doi:10.5604/01.3001.0055.9903
BibTeX
@article{Florek2026,
author = {Florek, Szymon and Górny, Piotr},
title = {Ablation study of correction methods in Polish language Retrieval-Augmented Generation chatbots},
journal = {Computer Science and Mathematical Modelling},
year = {2026},
volume = {0},
number = {23/2026},
pages = {5--18},
doi = {10.5604/01.3001.0055.9903}
}
EUVIMED