小红书 AllSpark 开源 Search Agent 模型 Iris，35B 与 397B 版本同量级成绩领先
AIHOT ·
小红书 AllSpark 团队发布并开源 Search Agent 模型 Iris，权重和评测代码已公开，数据与训练配方将陆续公布。
原文内容 · 暂无该语言译文AI News · 2026-09-14
按本站收录日期（UTC）归档，不代表事件发生于当天；原公告、聚合源发布与收录时间分别标注。
返回资讯列表LatentMD: Benchmarking Markdown Boundary Failures in LLM-Generated Text
arXiv cs.CL ·
Large language models (LLMs) increasingly generate Markdown that is consumed by renderers, agents, code extractors, and structured downstream pipelines. Yet existing evaluations…
原文内容 · 暂无该语言译文Do LLMs Trust the Accuser or the Accusation? Measuring Belief Shifts in Werewolf
arXiv cs.CL ·
Social-deduction games such as Werewolf are increasingly used to evaluate LLM agents, but existing evaluations often rely on final game outcomes. We propose a belief-shift evalu…
原文内容 · 暂无该语言译文HuRo: Robotizing Human Videos for Scalable VLA Pretraining
arXiv cs.RO ·
Human video datasets have emerged as a compelling alternative to expensive real-robot data, offering rich diversity at scale. To bridge the human-to-robot embodiment gap, existi…
原文内容 · 暂无该语言译文DIA: Denoising Intermediate Advantage for Diffusion Policy Optimization
arXiv cs.RO ·
Diffusion-based robot policies have become widely used in robotic manipulation, where they are typically trained with behavior cloning. However, policies trained purely from dem…
原文内容 · 暂无该语言译文Agent 长任务上下文工程解析：用预算控制、压缩、todo-state 和记忆对抗上下文溢出与目标丢失
AIHOT ·
文章解析 Agent harness 层应对长任务中上下文溢出与目标丢失的四类机制：上下文预算与卸载、压缩、todo-state 复述和跨会话记忆。
原文内容 · 暂无该语言译文LlamaIndex 解析 just-in-time Agentic OCR：两遍式文档处理如何平衡成本与精度
AIHOT ·
LlamaIndex 博客提出 just-in-time Agentic OCR 模式：先用 LiteParse 等免费解析器粗读全部文件供检索，再仅对相关页面调用 VLM 做 OCR。
原文内容 · 暂无该语言译文Hugging Face 用 LoRA 与 Storage Bucket 在 HF Jobs 上跑异步 GRPO，免 NCCL 并提速 3.9 倍
AIHOT ·
Hugging Face 在 TRL v1.14 的 AsyncGRPOTrainer 中支持只训练和同步 LoRA adapter，配合 Storage Bucket 挂载与代理路由，让训练 Job 和 vLLM 推理 Job 在不同机器上运行而无需 NCCL。
原文内容 · 暂无该语言译文NVIDIA PAIR Virtual Inference Router Expands Available Compute on Your Local Network
NVIDIA Technical Blog ·
AI agents are learning to do more by working together. A lead agent can break a complex task into smaller jobs and assign those jobs to specialized subagents.... AI agents are l…
原文内容 · 暂无该语言译文Co-Designing AI Models Using Speculative Decoding for Faster LLM Inference
NVIDIA Technical Blog ·
This post is the third in a series on AI model co-design. It explores how to accelerate LLM inference while maintaining accuracy using speculative decoding and... This post is t…
原文内容 · 暂无该语言译文