technique
Multi-hop Question Answering
techniqueactiveprovisional
multi-hop-question-answering-b3474c93·1 events·first seen 16d agoAliases: Multi-hop Question Answering
Co-occurring entities
More like this (12)
multi-hop reasoningMulti-hop Graph Retrieval2WikiMultiHopQABias Benchmark for Question AnsweringVisual Question AnsweringTrace Only What You Need: Structure-Aware On-Demand Hypergraph Memory for Long-Document Question AnsweringDocument Visual Question AnsweringTable Question Answeringtabular question answeringGoogle-Proof Q&AWhere Does the Answer Come From? Benchmarking View-Level Visual Evidence Identification in Multi-View MLLMs for Autonomous DrivingObservation-Guided Video-Context Routing
Recent events (1)
LongTraceRL: Reinforcement Learning for Long-Context Reasoning via Search Agent Trajectories and Rubric Rewards
LongTraceRL is a new RL training framework for improving long-context reasoning in LLMs, addressing limitations of existing RLVR methods. It constructs challenging training data using multi-hop questions from knowledge graph random walks and tiered distractors derived from search agent trajectories (high-confusability: read but uncited; low-confusability: seen but unopened). A rubric reward provides entity-level process supervision along reasoning chains, applied only to correct responses to prevent reward hacking. Experiments across three LLMs (4B–30B parameters) on five long-context benchmarks show consistent improvements over strong baselines.