Pai 的 AI 雷达
2026-08-13：The builder’s guide to GPT‑5；Previewing Ultrafast mode: G；OpenAI appoints Dali Rajic a
Recent work on distillation transfers the capabilities of large models to smaller ones often by updating the latter's parameters, through teacher forcing, on-policy distillation, and related training-time methods. In thi
The rapid advancement of Large Language Models (LLMs) is revolutionizing AI for Games by enabling open-ended and fluid interactive storytelling. However, existing research has largely overlooked the critical challenge of
Embodied agents are increasingly built as systems around foundation models, where performance depends not only on model weights but also on the skills, context, action interfaces, and execution harness surrounding the mo
Complex software systems develop over timescales that exceed the lifespan of any individual coding agent. Most agentic software systems preserve continuity through persistent sessions, memories, managers or shared contex
Agentic systems powered by large language models (LLMs) have opened new opportunities for business ideation. Yet existing approaches remain confined to a text-only paradigm, despite the inherently multimodal nature of re
Simulating societies of many large language model (LLM) agents is expensive, yet the questions asked of such simulations are usually macroscopic: phase behaviour, stylised facts, and scaling with the number of agents N,
Turning a research idea into a complete paper requires more than text generation: the system must retrieve literature, design and execute experiments, revise claims according to evidence, produce publication-ready figure
AI models have achieved remarkable success across diverse domains, yet the mechanisms underlying their capabilities and the risks they may pose remain poorly understood. As AI development becomes faster and increasingly
Large Language Models (LLMs) increasingly act as agents whose procedural knowledge is stored in reusable skill packages and loaded at inference time. As skill libraries grow, a central challenge is to expose the smallest
Previsualization is an intermediate layer between ideas and production in film, games, architecture, and urban design. It lets creators iteratively refine scenes, actions, cameras, and spatial-temporal dynamics. Yet exis
Recent Vision Foundation Models (VFMs) predict depth, camera pose, and pointmap in a single forward pass without per-scene optimization, achieving strong generalization. However, enforcing explicit multi-view geometric c
The world evolves following its dynamics, i.e., its laws of motion. However, leading video diffusion models largely fit the pixels without modeling how the pixels transit over time. Thus, they render visually plausible f
Large language model (LLM) agents integrated with external tools are vulnerable to indirect prompt injections embedded in environmental states. However, existing studies largely rely on manually implemented or reused env
Videos captured through glass often contain reflections that degrade visual quality and interfere with downstream vision tasks. Although single-image reflection removal has been extensively studied, video reflection remo
The "thinking-with-images" paradigm equips multimodal LLMs with active visual operations such as crop-and-zoom. However, models using these operations often achieve only marginal or negative gains over direct inference a
Discrete diffusion models for categorical generation are defined by a corruption kernel, which determines the intermediate state space and the associated reverse prediction problem. We study uniform discrete diffusion an
Estimating human gaze targets from images in-the-wild is an important and formidable task. Existing approaches primarily employ brittle, multi-stage pipelines that require explicit inputs, like head bounding boxes and hu
The dominant paradigm treats AI safety as a property to be instilled during model training via RLHF, DPO, or Constitutional AI. We argue this is structurally insufficient for autonomous agents that execute code, mutate f
ReRound (Reconstructive Rounding) is a post-training quantization method that addresses the midpoint ambiguity inherent in standard round-to-nearest (RTN) schemes when quantizing weights near the centers of quantization
Multimodal expansion of large language models (LLMs) enables new perceptual capabilities but often compromises the language intelligence acquired during pretraining. In this work, we investigate this phenomenon from the
While Vision-Language-Action (VLA) models have advanced embodied AI, their fundamentally reactive paradigm severely limits performance in partially observable and long-horizon tasks. When restricted to a single wrist-mou
World modeling is an unsettled field: architectures, training objectives, and state representations interact in complex ways, and no single recipe dominates across environments. This makes it an ideal testbed for AI codi
AI agents operate in persistent environments where early state changes can influence decisions far into the future. Unlike conventional language-model interactions, agent behavior is mediated through a shared state that
Exploration has been a focus of reinforcement learning research for a long time. Recently, there has been growing evidence that it is also an important ingredient in LLM reinforcement learning recipes that can significan