1
Context Tuning for In-Context Optimization
ICML 2026最新论文，提出上下文调优方法，让大模型在上下文优化中表现更高效。
arXiv:2507.04221v3 Announce Type: replace-cross Abstract: We introduce Context Tuning, a simple and effective method to significantly enhance few-shot…
2
ReContext: Recursive Evidence Replay as LLM Harness for Long-Context Reasoning
长上下文推理新突破，递归证据回放让LLM精准捕捉关键信息
arXiv:2607.02509v1 Announce Type: new Abstract: Understanding and reasoning over long contexts has become a key requirement for deploying large langua…
3
Show HN: theta_py, bindings to the Theta CLI
为Theta CLI提供Python绑定，让你在代码中直接调用AI命令行工具，提升开发效率。
Article URL: https://github.com/tamarillo-ai/theta_py Comments URL: https://news.ycombinator.com/item?id=48514641 Points: 1 # Comments: 0
4
In-Context Graphical Inference
新的上下文图推理范式，无需参数更新即可在图中实现快速预测与结构归纳。
arXiv:2606.05042v1 Announce Type: cross Abstract: Marginal inference in discrete graphical models forces a choice between exactness and scalability: e…
5
Learning without training: The implicit dynamics of in-context learning
不做梯度更新也能学习？这篇论文揭示了上下文学习背后的隐式动力学机制
arXiv:2507.16003v4 Announce Type: replace Abstract: One of the most striking features of Large Language Models (LLMs) is their ability to learn in-con…
6
Best MCP Servers for AI Coding in 2026: The 7 Worth Installing in Cursor, Claude Code, and Windsurf
2026年AI编码必装的7个MCP服务器，Context7+GitHub MCP组合覆盖90%工作流，Cursor/Claude Code/Windsurf用户必看。
This article was originally published on aicoderscope.com The verdict up front: Context7 and GitHub MCP belong in every setup. Add Playwright if you w…
7
Toward Robust In-Context Learning: Leveraging Out-of-distribution Proxies for Target Inaccessible Demonstration Retrieval
ACL 2026新研究，用分布外数据代理解决目标不可访问时的上下文示例检索，显著增强大模型ICL鲁棒性。
arXiv:2606.00014v1 Announce Type: cross Abstract: Although studies have demonstrated that Large Language Models (LLMs) can perform well on Out-of-Dist…
8
Prompting styles - Basic
一文快速掌握五种基础提示风格，零基础也能精准操控大模型输出。
Query which we ask the LLM is referred to as prompt. The way in which we provide prompt to LLM makes a difference and there are different ways to to p…
9
Best MCP Memory Servers for Teams in 2026: Context Cloud vs mem0 vs Basic Memory vs claude-mem vs MemPalace
2026年团队级MCP内存服务器横向评测，五大方案（Context Cloud、mem0等）优劣势全解析，助你选对AI记忆方案。
There are now dozens of MCP memory servers. I've tried most of them. They almost all solve the same problem: your AI forgets everything between sessio…
10
CurveRL: Principled Distribution-Aware Context Reweighting for LLM Reasoning
提出CurveRL方法，通过分布感知的上下文重加权提升LLM推理性能，理论严谨效果显著。
arXiv:2605.24331v1 Announce Type: new Abstract: Context or prompt-level reweighting has emerged as a central algorithmic lever in Reinforcement Learni…
11
Context-CoT: Enhancing Context Learning via High-Quality Reasoning Synthesis
Context-CoT通过高质量推理合成增强上下文学习，为AI大模型提供更优的示例生成策略。
arXiv:2605.25354v1 Announce Type: new Abstract: While LLMs excel at reasoning over prompts using static pretrained knowledge, they struggle significan…
12
MEMOR-E: In-Context and Fine-Tuned LLM Personalization for Alzheimer's Assistive Robotics
为阿尔茨海默症辅助机器人量身定制，融合上下文学习与微调的LLM个性化方案
arXiv:2605.23941v1 Announce Type: new Abstract: Alzheimer's disease is a neurodegenerative disorder marked by progressive declines in memory and langu…
13
Context: Proactive Goal-Directed Intelligence via Composable Sandboxed Programs, Declarative Wiring, and Structured Interaction
提出Context智能层架构，通过沙盒程序、声明式连接与结构化交互实现主动目标导向智能。
arXiv:2605.23928v1 Announce Type: new Abstract: We present Context, the intelligence layer of the Magarshak Architecture, which replaces reactive quer…
14
Fine-tuning vs. In-context Learning in Large Language Models: A Formal Language Learning Perspective
从形式语言学习视角，剖析微调与上下文学习的根本差异，ACL 2026 重磅论文。
arXiv:2604.23267v2 Announce Type: replace-cross Abstract: Large language models (LLMs) operate in two fundamental learning modes - fine-tuning (FT) an…
15
ACIL: Auto Chain of Thoughts for In-Context Learning
提出ACIL方法，自动化链式思维推理与上下文学习结合，提升LLM多步推理能力。
arXiv:2605.17088v1 Announce Type: new Abstract: Recent advances in large language models (LLMs) have shown that Chain-of-Thought (CoT) reasoning can s…
16
Correctness-Aware Repository Filtering Under Maximum Effective Context Window Constraints
当LLM的上下文窗口超过有效限制时，准确率骤降。这篇论文提出的正确性感知预过滤框架，无需索引、亚毫秒级别过滤，却能压缩近90% token、提升准确率至72%，击中了LLM代码工具的效率要害。
arXiv:2605.14362v1 Announce Type: cross Abstract: Context window efficiency is a practical constraint in large language model (LLM)-based developer to…