2026-08-18（周二） 收录 10 篇论文：HarnessEval-W: Agentifying the Evaluation of Visual Worlds；ClawGym II: Exploring Black-Box RL on Agent Harness；GenRouter: Unified Workflow Routing for Agentic Image Generation…
A benchmark should deliver more than a scalar score: what makes an evaluation trustworthy is the reasoning that justifies the score. This is especially critical for world models, where judging a rollo...
Agent harnesses have substantially improved performance on long-horizon tasks by coordinating agent interactions with the environment. However, reinforcement learning through complex harnesses remains...
The rapid evolution of text-to-image (T2I) generation models has effectively solved the foundational challenge of raw pixel synthesis, shifting the community's focus toward fulfilling increasingly int...
Agents now write knowledge graphs, but knowledge-graph stores still carry defaults set when humans curated them: accept writes now and clean later, keep one time axis or none, treat every writer's fac...
Autonomous agents share a transport and can call each other's tools, but they cannot share what they know: no protocol lets two agents' memories reconcile a fact phrased two ways, link related facts h...
LLM-based agents can act on behalf of a user to access cloud services, call tools, or invoke agents. At session start, the agent's permissions are set but remain static, and each request is evaluated ...
In cognitive science, resource rationality asks how an agent should allocate limited computation to maximize expected value. Most reasoning and agent benchmarks use independent per-task budgets; exist...
This paper investigates an increasingly important topic in generative modeling: pixel-space diffusion models. Although numerous studies have explored this topic, most focus on small-scale or class-con...
We present AnyTalk, a novel method for generating 3D speech animations for arbitrary characters without requiring any animation data. While existing audio-driven 3D speech animation methods rely on ch...
Automated checking pipelines increasingly place one language model as the checker and another (or the same one) as the fixer. We ask whether that wiring changes what the checker reports. Measuring fal...