Multi-Modal29 [2026-1] 백승우 - Agentic Reward Modeling: Verifying GUI Agent via Online Proactive Interaction Agentic Reward Modeling: Verifying GUI Agent via Online Proactive InteractionReinforcement learning with verifiable rewards (RLVR) is pivotal for the continuous evolution of GUI agents, yet existing evaluation paradigms face significant limitations. Rule-based methods suffer from poor scalability and cannot handle open-ended tasks,arxiv.org 2026. 3. 24. [2026-1] 강민정, 염제원 - GDPval: Evaluating AI Model Performance on Real-World Economically Valuable Tasks Paperhttps://arxiv.org/abs/2510.04374 GDPval: Evaluating AI Model Performance on Real-World Economically Valuable TasksWe introduce GDPval, a benchmark evaluating AI model capabilities on real-world economically valuable tasks. GDPval covers the majority of U.S. Bureau of Labor Statistics Work Activities for 44 occupations across the top 9 sectors contributing to U.S. GDParxiv.orgArticlehttps://.. 2026. 3. 20. [2026-1] 백승우 - AutoWebWorld: Synthesizing Infinite Verifiable Web Environments via Finite State Machines https://arxiv.org/abs/2602.14296 AutoWebWorld: Synthesizing Infinite Verifiable Web Environments via Finite State MachinesThe performance of autonomous Web GUI agents heavily relies on the quality and quantity of their training data. However, a fundamental bottleneck persists: collecting interaction trajectories from real-world websites is expensive and difficult to verify. Tarxiv.org 2026. 3. 10. [2026-1] 정유림 - FiLM: Visual Reasoning with a General Conditioning Layer paper link : https://arxiv.org/pdf/1709.07871 CLEVR datset : 다단계 추론의 학습이 필요. 기존 방법의 성능이 좋지않았음.reasoninng 능력 평가 : CLEVR datset : 다단계 추론의 학습이 필요. 기존 방법의 성능이 좋지않았음.FiLM (Feature-wise Linear Modulation): 조건 입력(질문)에 따라, 신경망 중간 feature에 대해, feature별 변환 수행. 시각적 추론에서, FiLM layer를 추가해서, 질문을 처리하는 RNN이 이미지 처리를 담당하는 CNN의 계산에 영향을 미치게됨.즉, 질문의 내용에 따라 이미지를 처리하는 방식 자체가 달라짐.→ Conditional Normalization의 일반화로 볼수있.. 2026. 2. 21. 이전 1 2 3 4 ··· 8 다음
