2026年9月17日 技术热点总结

项目简介：A mascot that watches the cursor and blinks when you poke it

技术栈：Python

核心介绍：An interactive character that watches the cursor and blinks when you poke it.

项目数据：⭐ 629 Stars，🍴 55 Forks

项目简介：大模型（LLM）全栈学习路线与中文教程🔥：覆盖 Prompt Engineering、RAG、AI Agent、MCP、微调、模型部署、Transformer、AI 编程与大厂面试，从入门到生产实践。

技术栈：agent、large-language-models、llm、llm-learning、llm-tutorial、rag

核心介绍：> 本仓库资料来自卡码笔记-大模型专栏。 从第一次调用模型 API，到构建可评估的 RAG、可靠的 Agent 与可上线的 AI 系统。 开始学习 · 全部教程 · 专题索引 · 面试题库 网上从来不缺大模型资料，缺的是一条真正适合程序员的学习路径。 LLM Master 不是零散文章的合集，而是一套从基础认知 → 应用开发 → RAG → Agent → 微调与部署 → Transformer → 项目与面试逐级展开的完整知识体系。它关心的不只是“这个概念是什么”，更关心：为什么这样设计、工程上如何落地、出了问题怎么排查、面试时如何讲清楚。 如果你有 Java、C++、Go、Python 或前端开发基础，希望进入大模型应用开发、Agent 工程或 AI 后端，这个仓库就是为你准备的。 A[0 · 全局认知] –> B[1 · 模型调用] D –> E[4 · 生产工程]

**项目数据：…

Masked diffusion language models (dLLMs) generate text by iteratively denoising masked tokens with bidirectional attention. Extending reasoning across generation chunks normally requires keeping earlier generated text in context. We ask whether a dLLM can instead continue reasoning after that text is cleared, using only a fixed-size carried state. We implement this state as a small number of register tokens: dedicated fixed-position tokens whose continuous hidden states are trained to carry reasoning progress across generation chunks. We post-train dLLMs to decode a chunk of text, clear it …

Traditional multimodal representation learning and generation are two stages: a contrastive or self-supervised visual encoder is trained first, followed by a separate downstream generative model. This setup bottlenecks generative performance behind frozen embeddings. To bridge this gap, we revisit joint multimodal representation learning and generation to produce linearly interpolatable embeddings that are directly consumable by generative decoders. We present FLAT (Flexible-Length Aligned Transmodal representations), a representation pre-training framework that jointly optimizes a shared m…

Open-vocabulary detection accepts any class list at inference, and promptable segmentation returns regions without class names: the taxonomy has left the model and become an input. Relation prediction has not. Scene-graph models are still trained and evaluated on the 50 or 56 predicates of one annotation style, their relation head conditioned on object labels and so tied to one detector. Three obstacles explain this, none primarily modelling: no relation corpus is both free-text and verified, a label-conditioned architecture cannot accept a vocabulary it was not trained on, and the standard…

When we speak of recursive self-improvement (RSI), are we speaking of a phenomenon, a mechanism, or a prospect? Towards autonomous and evolving intelligence, RSI is being claimed at many scales, while no single framework that formally describes these emerging instances exists. Its counterpart in the classical realm, iterative policy improvement, is characterized by generalized policy iteration (GPI), a framework of broad applicability with well-understood theoretical properties, but only where the update principle and the evaluation base lie outside the agent. In this paper, we propose Gene…

Unified multimodal large language models (MLLMs) and multi-agent systems have advanced visual generation. However, three limitations remain. (1) Existing methods often distill task-specific experience with limited generalizability. (2) Reflection is often deferred until task completion. (3) Knowledge is often acquired only in response to downstream task demands. To address these limitations, we introduce OmniHarness, a framework for generalizable visual generation via symbolic policy learning. OmniHarness abstracts verified executions into symbolic policies for visual generation task famili…

Language model-generated rubrics are increasingly used as reward signals for rubric-based reinforcement learning, LLM-as-a-judge evaluation, and automated grading. Such rubrics are reliable only if they reward honest answers over adversarial answers optimized to exploit them. Yet their robustness to such optimization remains poorly understood. We isolate the hardest regime: impossible tasks, where the prompt pressures the model toward an unsupported conclusion, so the only honest response is to acknowledge the impossibility. We introduce ImpossibleRubrics, a benchmark of 169 impossible task…