[논문리뷰] AgentLongBench: A Controllable Long Benchmark For Long-Contexts Agents via Environment RolloutsarXiv에 게시된 'AgentLongBench: A Controllable Long Benchmark For Long-Contexts Agents via Environment Rollouts' 논문에 대한 자세한 리뷰입니다.#Review#Long-Context LLMs#Autonomous Agents#Benchmark#Environment Rollouts#State Tracking#Tool Use#Memory Evaluation#Lateral Thinking Puzzles2026년 1월 29일댓글 수 로딩 중