1
2
Versiona acciones de correo en agentes LLM
用版本号锁定邮件动作，让LLM agent不再自由发挥，提升可预测性与可靠性。
Muchos equipos que integran LLMs con correo se obsesionan con el prompt y dejan medio borroso el contrato de ejecución. En mi experiencia, el fallo re…
3
Steerable Visual Representations
最新ECCV 2026录用论文，探索如何对预训练视觉Transformer的表示进行可控操作，为视觉AI带来新的灵活性。
arXiv:2604.02327v2 Announce Type: replace Abstract: Pretrained Vision Transformers (ViTs) such as DINOv2 and MAE provide generic image features that c…
4
Building Controllable AI Agents: Why I Stopped Using Black-Box Tool-Spammers
黑盒工具不可控？终端原生AI执行环境Deepstrain用反脆弱架构让代理可控可观测。
The Problem with Most AI Coding Agents If you've tried AutoGPT, CrewAI, or similar agent frameworks, you've probably seen this: the agent starts spamm…
5
ChatSOP: An SOP-Guided MCTS Planning Framework for Controllable LLM Dialogue Agents
提出SOP引导的MCTS规划框架，让LLM对话代理在复杂任务中更可控、更高效。
arXiv:2407.03884v4 Announce Type: replace-cross Abstract: Dialogue agents powered by Large Language Models (LLMs) show superior performance in various…
6
Agentic Chain-of-Thought Steering for Efficient and Controllable LLM Reasoning
新方法通过智能体引导链式思维，让大模型推理更高效、更可控。
arXiv:2606.03965v1 Announce Type: cross Abstract: Large language models improve final-answer accuracy through extended chain-of-thought reasoning, but…
7
Agentic Clustering: Controllable Text Taxonomies via Multi-Agent Refinement
多智能体协作的文本聚类新范式，让分类更可控且精准。
arXiv:2606.01255v1 Announce Type: new Abstract: Recent text-clustering methods use large language models to propose a cluster taxonomy from a corpus a…
8
Ask HN: Is anybody providing deterministic LLMs?
探讨大语言模型确定性输出的可行性与现状，引发对AI可靠性的思考
Comments URL: https://news.ycombinator.com/item?id=48325265 Points: 4 # Comments: 8
9
Boundary Suppression Asymmetry in Post-trained Assistants: Over-expansion as a Controllability Cost
大模型后训练中的边界抑制不对称，揭示过度扩展是可控性代价——来自前沿研究的深度洞察。
arXiv:2605.27969v1 Announce Type: new Abstract: Post-trained language-model assistants are often optimized to avoid under-answering, encouraging compl…
10
Position: AI Safety Requires Effective Controllability
AI安全领域重磅论文，提出有效可控性是实现AI安全的必要条件
arXiv:2605.27117v1 Announce Type: new Abstract: AI safety is still largely framed as alignment: training models to follow human preferences, safety po…
11
Controlla: Learning Controllability via Graph-Constrained Latent Geometry
新方法Controlla通过图约束潜在几何实现可控性学习，为生成模型中的控制能力提供理论框架。
arXiv:2605.16603v1 Announce Type: new Abstract: Controllable multimodal generation is commonly formulated as an inference-time conditioning problem us…
12
Activation Steering with a Feedback Controller
用反馈控制器实现精准激活引导，为提升大模型可控性提供新思路，ICLR2026论文。
arXiv:2510.04309v3 Announce Type: replace Abstract: Controlling the behaviors of large language models (LLM) is fundamental to their safety alignment …