3 papers
cs.CL2026
ToolMATH: A Diagnostic Benchmark for Long-Horizon Tool Use under Systematic Tool-Catalog Constraints
Hyeonje Choi, Jeongsoo Lee, Hyojun Lee +1
We introduce \ToolMATH, a math-grounded diagnostic benchmark for evaluating long-horizon tool use under controllable tool-catalog conditions. \ToolMATH converts stepwise MATH solut…
cs.CL2025
GRADE: Generating multi-hop QA and fine-gRAined Difficulty matrix for RAG Evaluation
Jeongsoo Lee, Daeyong Kwon, Kyohoon Jin
Retrieval-Augmented Generation (RAG) systems are widely adopted in knowledge-intensive NLP tasks, but current evaluations often overlook the structural complexity and multi-step re…
cs.IR2025
MHTS: Multi-Hop Tree Structure Framework for Generating Difficulty-Controllable QA Datasets for RAG Evaluation
Jeongsoo Lee, Daeyong Kwon, Kyohoon Jin +3
Existing RAG benchmarks often overlook query difficulty, leading to inflated performance on simpler questions and unreliable evaluations. A robust benchmark dataset must satisfy th…