Do LLMs Trust the Accuser or the Accusation? Measuring Belief Shifts in Werewolf Benchmark

What is Do LLMs Trust the Accuser or the Accusation? Measuring Belief Shifts in Werewolf?

A proposed belief-shift evaluation in Werewolf for analyzing communication skills through belief updating. Benchmark Radar tracks Do LLMs Trust the Accuser or…

Released

2026-09-11

Evaluates

General AI, Language & Knowledge, cs.AI

Openness

unknown

Importer

Claire Radar

Review status

ai-name-audit-deferred

Reported scores

0

Source provenance

• Original evidence https://arxiv.org/abs/2609.12446

Do LLMs Trust the Accuser or the Accusation? Measuring Belief Shifts in Werewolf paper

• Paper https://arxiv.org/abs/2609.12446
• Website https://rlg.iis.sinica.edu.tw/papers/werewolf-accusation-benchmark

Which sources cite Do LLMs Trust the Accuser or the Accusation? Measuring Belief Shifts in Werewolf?

• Do LLMs Trust the Accuser or the Accusation? Measuring Belief Shifts in Werewolf

Related General AI, Language & Knowledge, cs.AI benchmarks

• $\tau^\tau$-Bench: An Environment for End-To-End, Realistic Agent Construction
• 3DCodeBench
• 5-Dialects-BN
• A Large-Scale Dataset and Benchmark
• A3S-Bench
• AARRI-Bench
• ABC-Bench
• ActBench