Comment history outperforms explicit personas for simulating human discussion behavior

A multilingual benchmark from Wikipedia Talk pages reveals systematic biases toward agreeableness in LLM agents

Big Tech
Dennis Fucci · Andrea Bacciu · Dong Liu · Weronika Łajewska · Saab Mansour

Amazon

Research Digest··2 min read
The authors introduce Wiki-Talkie, a multilingual dataset of real Wikipedia discussions with ground-truth user personas across five languages.

The authors constructed Wiki-Talkie from Wikipedia Talk pages in English, German, Spanish, French, and Italian, spanning two language families.

Why this paper

From Amazon

In one line

Comment history outperforms explicit persona descriptions for simulating human discussion patterns in multilingual agent interactions.

What we could check

  • ·No code link found
  • ·No weights link found
  • ·No dataset link found
  • ·No compute details found
  • ✓Limitations stated by the authors
  • ·No benchmark numbers found

Observed from the paper text and links we have. Absence here means we did not find it, not that it does not exist.

§

Research Digest

Written by software from the reporting listed above, scored by an automated standards desk, and published without a person reading it first. If something here is wrong, tell the editor and it will be put right.