Tag
Wizard 提出一种游戏世界模型，通过解耦角色动画、场景操控与游戏逻辑，实现对未见过角色的关节级动作与实时交互控制，支持 24 FPS 交互并对新场景、角色和动作具备零样本泛化能力。
The paper introduces Stateful Harness, a framework that separates rule-based world state updates from frozen video rendering to improve consistency in interactive world models with hidden state, and evaluates it with the WorldState-Bench benchmark, where the full harness raises rule-compliance from 32% to 92%.
本文提出MIMESIS，一个基于真实用户对话与显式推理监督训练的专用用户模拟器（9B模型），在行为保真度上超越前沿模型；研究还将该模拟器冻结后用于多轮强化学习训练智能体，并提出Coached On-Policy Self-Distillation（CSD）方法，在九个未见用户模拟器上实现更强泛化性能。
KlinikeBench introduces a clinician-verified benchmark of 333 interactive clinical tasks where LMs must gather patient history, use clinical tools, and reach a diagnosis within a turn budget. Across 31 models, the best (e.g., GPT-6-astra, Claude Opus 5) succeed on under 30% of tasks despite 90.7% diagnostic accuracy, revealing a large gap between diagnostic knowledge and real-world clinical interaction.
This paper introduces causal retention in interactive agents, examining how frozen learned states can answer intervention queries independently of training objectives, supported by theoretical analysis and experiments on finite causal systems and AI models like Qwen.
ScienceBuddy introduces a recursive-in-recursive self-improvement paradigm for interactive scientific agents, enabling continual evolution through researcher collaboration and feedback.
VoiceChat-TTS is a low-latency, continuous text-to-speech model designed for interactive agents, enabling real-time streaming and interruption handling without compromising speech quality.
Proposes Online Agent-as-a-Judge, an evaluation framework that uses an in-world evaluator agent to actively generate situations for testing interactive social agents, improving coverage and reliability over passive methods.
The paper proposes the Map-then-Act Paradigm (MAP), a plug-and-play framework that shifts environmental understanding before execution in interactive LLM agents, achieving consistent gains across benchmarks and enabling frontier models to surpass near-zero baseline performance in 22 of 25 game environments.