REALHOP: Rethinking Multi-Hop Reasoning Evaluation via Behavioral Auditing

arXiv:2609.36984 · cs.AI · Submitted 2026-09-29 · Read on arXiv

cs.AI

Submitted: 2026-09-29

Updated: 2026-09-29

Project page: https://loongrl.github.io

Terminology

• Multi-hop Reasoning — 1× in this paper · explained in 3 episodes

Sources

• BABILong: Testing the Limits of LLMs with Long Context Reasoning-in-a-Haystack
• Measuring Faithfulness in Chain-of-Thought Reasoning
• LongTraceRL: Learning Long-Context Reasoning from Search Agent Trajectories with Rubric Rewards
• MoreHopQA: More Than Multi-hop Reasoning
• Beyond Rephrasing: Book-Level Organization Improves Synthetic Textbook Data for Mid-Training
• Language Models Don't Always Say What They Think: Unfaithful Explanations in Chain-of-Thought Prompting

Related papers

• MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
• Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
• Agents' Last Exam
• The Clinician's Veto: Navigating Trust, Liability, and Uncertainty in Autonomous AI Prescribing
• MindHelper: Closed-Loop Embodied Mental-State Reasoning for Precision Intervention
• Incumbent Advantage: Brand Bias and Cognitive Manipulation Dynamics in LLM Recommendation Systems