English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Zendy Plus

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Zendy Tools

Zendy Open

Multi-agent debate system (MAD) imitating the process of human discussion inpursuit of truth, aims to align the correct cognition of different agents forthe optimal solution. It is challenging to make various agents perform rightand highly consistent cognition due to their limited and different knowledgebackgrounds (i.e., cognitive islands), which hinders the search for the optimalsolution. To address the challenge, we propose a novel\underline{M}ulti-\underline{A}gent \underline{D}ebate with\underline{K}nowledge-\underline{E}nhanced framework (\textbf{MADKE}) topromote the system to find the solution. First, we involve a shared retrievalknowledge pool in the debate process to solve the problem of limited anddifferent knowledge backgrounds. Then, we propose an adaptive knowledgeselection method to guarantee the accuracy and personalization of knowledge.This method allows agents to choose whether to use external knowledge in eachconversation round according to their own needs. Our experimental results onsix datasets show that our method achieves state-of-the-art results compared toexisting single-agent and multi-agent methods. Further analysis reveals thatthe introduction of retrieval knowledge can help the agent to break cognitiveislands in the debate process and effectively improve the consistency andcorrectness of the model. Moreover, MADKE using Qwen1.5-72B-Chat surpassesGPT-4 by +1.26\% on average in six datasets, which validates that our methodcan help open-source LLMs achieve or even surpass the performance of GPT-4. Ourcode is available at \url{https://github.com/FutureForMe/MADKE}.

Learning to Break: Knowledge-Enhanced Reasoning in Multi-Agent Debate System