发布日期：2026-08-15
收录条目：20
1. Mark Zuckerberg has an Instagzam
- 来源：The Verge AI
- 发布时间：2026-08-14 16:54 UTC
- 链接：https://www.theverge.com/podcast/980367/instagram-logo-new-zuckerberg-ai-vergecast
摘要：Instagram's wordmark is iconic. Well, was iconic. Apparently Instagram thought it looked old, so the company rolled out a new one this week. It doesn't look like the old Instagram wordmark. It doesn't even look like it s
2. You can now turn off Google Gemini’s visible watermarks
- 来源：The Verge AI
- 发布时间：2026-08-14 16:39 UTC
- 链接：https://www.theverge.com/tech/980416/google-gemini-ai-watermarks-removal
摘要：Google will now allow you to remove visible watermarks from the images, videos, and music made with AI tools. With the update, you can toggle off a new "Media watermark" setting in Gemini and Google's AI video generator,
3. Custom reward functions for multi-turn reinforcement learning with Amazon Nova Forge
- 来源：AWS ML Blog
- 发布时间：2026-08-14 16:02 UTC
- 链接：https://aws.amazon.com/blogs/machine-learning/custom-reward-functions-for-multi-turn-reinforcement-learning-with-amazon-nova-forge/
摘要：In multi-turn reinforcement learning, your custom reward function decides what the model actually learns. This post shows how to design a composite multi-turn reward for Amazon Nova Forge, execute model-generated code sa
4. Building agentic workflows with SageMaker AI and Bedrock AgentCore
- 来源：AWS ML Blog
- 发布时间：2026-08-14 15:58 UTC
- 链接：https://aws.amazon.com/blogs/machine-learning/building-agentic-workflows-with-sagemaker-ai-and-bedrock-agentcore/
摘要：Learn how to combine OpenAI-compatible endpoints on Amazon SageMaker AI with Amazon Bedrock AgentCore runtime to build a multi-agent workflow where each specialized agent uses the model best suited to its job. This post
5. Apple trained its own AI model for China with help from Alibaba
- 来源：The Verge AI
- 发布时间：2026-08-14 09:21 UTC
- 链接：https://www.theverge.com/ai-artificial-intelligence/980160/apple-intelligence-china-custom-ai-model-alibaba
摘要：Apple has reportedly trained a custom AI model for the China market alongside domestic tech giant Alibaba, a rare cross-border partnership that cuts across growing tensions between Beijing and Washington. The China-focus
6. Z.ai Ships GLM-5.3 Without Retraining the Base Model: Better at Complex Coding and Long-Horizon Tasks
- 来源：MarkTechPost
- 发布时间：2026-08-14 08:03 UTC
- 链接：https://www.marktechpost.com/2026/08/14/z-ai-ships-glm-5-3-without-retraining-the-base-model-better-at-complex-coding-and-long-horizon-tasks/
摘要：Z.ai released GLM-5.3 on August 14, 2026. The model reuses the 743B GLM-5.2 base unchanged. Every reported gain comes from scaled post-training: more long-horizon task environments, more environment types, longer trainin
7. Meet Needle 2: An Open 45M-Parameter Tool-Calling Model That Ships as a 14MB Binary and Runs a Full Session in 28MB of RAM
- 来源：MarkTechPost
- 发布时间：2026-08-14 05:45 UTC
- 链接：https://www.marktechpost.com/2026/08/13/cactus-compute-needle-2-45m-parameter-tool-calling-model/
摘要：Cactus Compute released Needle 2, an open 45M-parameter model for tool calling, device use, and structured extraction. The full model is a single 14MB binary that runs a session in about 28MB of RAM. It leads both Seal-T
8. Position: Reasoning is a Learnable Rule-Based Process
- 来源：arXiv cs.AI
- 发布时间：2026-08-14 04:00 UTC
- 链接：https://arxiv.org/abs/2608.12325
摘要：arXiv:2608.12325v1 Announce Type: new Abstract: Autonomous reasoning is among the most scientifically and economically motivating topics in AI today. Historically the purview of symbolic AI, recent advances have mainly e
9. Diagnostic Foundation for Evaluating LLMs' Research Integrity as Co-Scientists
- 来源：arXiv cs.AI
- 发布时间：2026-08-14 04:00 UTC
- 链接：https://arxiv.org/abs/2608.12345
摘要：arXiv:2608.12345v1 Announce Type: new Abstract: Language models are increasingly deployed as co-scientists, yet their ability to uphold research integrity under institutional pressure remains unmeasured. We introduce Int
10. Position: The Alignment Community is Unintentionally Building a Censor's Toolkit
- 来源：arXiv cs.AI
- 发布时间：2026-08-14 04:00 UTC
- 链接：https://arxiv.org/abs/2608.12346
摘要：arXiv:2608.12346v1 Announce Type: new Abstract: This position paper argues that modern AI alignment methods - originally designed to prevent harmful output - are dual-use technologies that may easily be misused by malici
11. Agreement Is Not Alignment: Divergent Moral Grounds in Human and LLM Ethical Judgments
- 来源：arXiv cs.AI
- 发布时间：2026-08-14 04:00 UTC
- 链接：https://arxiv.org/abs/2608.12368
摘要：arXiv:2608.12368v1 Announce Type: new Abstract: Agreement with human judgments is a common proxy for evaluating the alignment of large language models (LLMs). Yet agreement in final labels does not show that human annota
12. Multi-Agent Scheduling with LLM-Assisted Contract Net Negotiation for Stream Processing in Mobile Edge Computing
- 来源：arXiv cs.AI
- 发布时间：2026-08-14 04:00 UTC
- 链接：https://arxiv.org/abs/2608.12371
摘要：arXiv:2608.12371v1 Announce Type: new Abstract: Stream-processing systems increasingly operate across heterogeneous mobile edge--cloud infrastructures, where workload volatility, resource contention, and stringent qualit
13. Position: We Need Practical AI Alignment Methods to Mirror Human Reasoning
- 来源：arXiv cs.AI
- 发布时间：2026-08-14 04:00 UTC
- 链接：https://arxiv.org/abs/2608.12372
摘要：arXiv:2608.12372v1 Announce Type: new Abstract: AI systems are increasingly employed as decision aids, decision delegates, or autonomous decision-makers. This position paper argues that in many settings, particularly hig
14. Don't Want Your LLM to Recommend Nuclear Strike? Try Asking It in Japanese
- 来源：arXiv cs.AI
- 发布时间：2026-08-14 04:00 UTC
- 链接：https://arxiv.org/abs/2608.12373
摘要：arXiv:2608.12373v1 Announce Type: new Abstract: Large language models are increasingly used in strategic and advisory contexts, yet their safety alignment is typically evaluated in English only. We test nine models from
15. Dual-Flow Transformers: Decoupling the Primary Prefill Path from Additional Decode Computation
- 来源：arXiv cs.AI
- 发布时间：2026-08-14 04:00 UTC
- 链接：https://arxiv.org/abs/2608.12385
摘要：arXiv:2608.12385v1 Announce Type: new Abstract: As large language models serve more requests, cumulative inference cost is becoming increasingly important relative to one-time training cost. The two inference phases stre
16. Learning to Adapt Cross-Domain Preferences via Meta-LoRA for LLM Personalization
- 来源：arXiv cs.AI
- 发布时间：2026-08-14 04:00 UTC
- 链接：https://arxiv.org/abs/2608.12389
摘要：arXiv:2608.12389v1 Announce Type: new Abstract: Cross-domain zero- or few-shot personalization aims to generate user-preferred responses in unseen conversational domains from only a handful of target-domain interactions.
17. Research Assistant: AstraZeneca's Agentic System for R&D
- 来源：arXiv cs.AI
- 发布时间：2026-08-14 04:00 UTC
- 链接：https://arxiv.org/abs/2608.12395
摘要：arXiv:2608.12395v1 Announce Type: new Abstract: We describe Research Assistant, an internal LLM-based system developed at AstraZeneca to help scientists and clinicians explore biomedical questions across a broad range of
18. Large Language Models Can Follow Instructions, But Not Many at Once: Phase Transitions in Compositional Constraint Satisfaction
- 来源：arXiv cs.AI
- 发布时间：2026-08-14 04:00 UTC
- 链接：https://arxiv.org/abs/2608.12426
摘要：arXiv:2608.12426v1 Announce Type: new Abstract: Large language models are increasingly deployed in settings that require simultaneous adherence to multiple explicit constraints - reasoning structure, safety boundaries, o
19. MindMemOS: A Portable and Self-Evolving Memory Operating Layer for AI Agents
- 来源：arXiv cs.AI
- 发布时间：2026-08-14 04:00 UTC
- 链接：https://arxiv.org/abs/2608.12428
摘要：arXiv:2608.12428v1 Announce Type: new Abstract: Memory is a core component of AI agents, enabling them to accumulate experience, maintain personalization, and adapt over long-term interactions. However, existing memory s
20. Governed Persistent Memory: Source-Bound State Semantics and Fail-Closed Release for Long-Horizon Agents
- 来源：arXiv cs.AI
- 发布时间：2026-08-14 04:00 UTC
- 链接：https://arxiv.org/abs/2608.12476
摘要：arXiv:2608.12476v1 Announce Type: new Abstract: Long-term agent memory is usually treated as select--store--retrieve, but retrieval does not decide whether contradictory, superseded, retracted, deleted, or stale records