1
GPT-Red: Unlocking Self-Improvement for Robustness
OpenAI用自我对弈实现自动化红队测试，大幅提升AI安全与提示注入防御能力。
Explore GPT-Red, OpenAI’s automated red teaming system that uses self-play to improve AI safety, alignment, and prompt injection robustness.
2
CANDI: Contextual Alignment for Niche Domains Question Answering
针对小众领域问答的上下文对齐新方法，提升大模型在专业场景下的精准回答能力。
arXiv:2607.11891v1 Announce Type: new Abstract: The deployment of large language models (LLMs) in specialized domains like medical diagnostics and fin…
3
When and Why Does Multi-Agent Debate Fail and Does It Really Underperform?
多智能体辩论为何失效？这篇论文系统分析失败条件，挑战你对AI协作的固有认知。
arXiv:2510.20963v2 Announce Type: replace Abstract: Multi-agent debate (MAD) was proposed as a promising approach for ensembling the wisdom of multipl…
4
How LLMs Learn to Be Helpful (RLHF vs DPO)
一文对比RLHF与DPO两种主流大模型训练方法的核心差异与适用场景
In this article, we will look at how that learning actually happens, starting with why instruction-following alone falls short, then walking through t…
5
Enhancing LLMs through human feedback: a journey towards self-improvement
从人类反馈到大模型自我进化，看最新研究成果如何用反馈驱动LLM性能跃升。
arXiv:2607.11267v1 Announce Type: cross Abstract: In the rapidly evolving landscape of information retrieval systems, the ability to adapt and improve…
6
Selective Safety Steering via Value-Filtered Decoding
提出选择性安全引导方法，通过价值过滤解码提升LLM安全性
arXiv:2605.14746v2 Announce Type: replace Abstract: While large language models (LLMs) are trained to align with human values, their generations may s…
7
Alignment Plausibility: A New Standard for Assuring AI in Healthcare
心理健康AI面临注意力经济挑战，新标准"对齐合理性"为医疗应用划出关键红线
arXiv:2607.07766v1 Announce Type: new Abstract: Large language models (LLMs) have become significant providers of mental health support, yet they rema…
8
Aligning Clinical Needs and AI Capabilities: A Survey on LLMs for Medical Reasoning
从临床需求出发，系统梳理LLM在医疗推理中的进展与对齐挑战。
arXiv:2607.07761v1 Announce Type: new Abstract: Large language models (LLMs) have emerged as important tools in healthcare, showing growing potential …
9
Aleena: Alignment Agent for Research Software Engineering Collaborations
专为研究软件工程协作设计的AI对齐代理Aleena，让AI队友更懂人类科学家的需求与节奏！
arXiv:2607.08043v1 Announce Type: cross Abstract: Research software collaborations span meetings, informal chats, pull requests, and GitHub issues. A …
10
Learning social norms enhances compatibility in dynamic human-AI coordination
让AI学会人类社会的潜规则，动态合作不再鸡同鸭讲
arXiv:2607.07021v1 Announce Type: new Abstract: Humans continuously coordinate with others in dynamic interactions, often through implicit, hard-to-qu…
11
Dissociating the Internal Representations of Sycophancy in LLMs
从神经层面拆解大模型「拍马屁」行为的内部机制，一篇被ICML 2026研讨会接收的机械可解释性突破。
arXiv:2607.07003v1 Announce Type: new Abstract: Large Language Models (LLMs) frequently exhibit sycophancy, where they agree with a user's statement e…
12
Refusal training for LLM agents against disguised MCP attacks
针对LLM代理的伪装MCP攻击，提出改进偏好对齐的拒绝训练方法，附RAG免训练对齐方案。
Article URL: https://github.com/johnhalloran321/mcp_safety_training Comments URL: https://news.ycombinator.com/item?id=48834106 Points: 1 # Comments: …
13
Specification Grounding Drives Test Effectiveness for LLM Code
揭示LLM代码测试：规范对齐才是提升有效性的关键，而不仅仅是自写测试。
arXiv:2607.06636v1 Announce Type: cross Abstract: Large language models frequently generate code that appears correct on typical inputs yet fails on e…
14
DT-Guard: Intent-Driven Reasoning-Active Training for Reasoning-Free LLM Safety Guardrail
无需额外推理即可强化大模型安全，意图驱动的新型护栏训练方法。
arXiv:2607.06326v1 Announce Type: new Abstract: Large language models deployed in open-world applications require safety guardrails that are both robu…
15
OpenAI’s Chief Futurist Is Leaving the Company
OpenAI首席未来学家Joshua Achiam离职，公司使命团队何去何从？
Joshua Achiam spent nearly nine years at OpenAI researching AI safety and made a memorable appearance in the Musk v. Altman trial.
16
What are They Thinking? Delineation, Probing, and Tracking of Concepts in LLMs
揭示LLM内部概念表征机制，为模型可解释性和安全对齐提供新方法。
arXiv:2605.28823v2 Announce Type: replace Abstract: As the influence of LLMs expands, it is imperative to gain insight into their decisions. One way t…
17
Trust-Region Noise Search for Black-Box Alignment of Diffusion and Flow Models
扩散与流模型的黑盒对齐新方法，信任区域噪声搜索带来高效微调
arXiv:2603.14504v2 Announce Type: replace-cross Abstract: Optimizing the noise samples of diffusion and flow models is an increasingly popular approac…
18
Not All Refusals Are Equal: How Safety Alignment Fails Cybersecurity at Scale
安全对齐虽关键，但一刀切的拒绝机制在网络安全等高风险领域适得其反，这篇论文揭示了系统性缺陷。
arXiv:2607.02714v1 Announce Type: cross Abstract: There is no doubt that safety alignment is an essential step in LLM training. However, conceptually …
19
Panoptes – AI audit and alignment layer
一键审计AI代理行为，追踪凭证访问与文件修改，让黑箱操作无所遁形
Article URL: https://github.com/miggy-code/Panoptes Comments URL: https://news.ycombinator.com/item?id=48813183 Points: 1 # Comments: 1
20
DPO vs RLHF: The Alignment Tax You Pay Without Knowing
对比DPO与RLHF的对齐代价，揭示大模型隐藏的哲学回答偏差
Ask yourself one question. When you talk to ChatGPT or Claude, do you feel like you talk to something that thinks — or something that agrees with you …