2026
ProofTeller: Exposing Recency Bias in LLM Reasoning and Its Side Effects on Communication (Extended Abstract)
ProofTeller, a benchmark that evaluates large language models' ability to faithfully interpret its reasoning, finds a consistent near-conclusion bias: LLMs tend to focus on steps closest to the final proof conclusion rather than on the most informative ones.
Mayank Jobanputra, A. Kovtunova, Brisca Balthes et al.
· Digital library · 0 citations