发布日期：2026-05-26
收录条目：20
1. Together AI Open-Sources OSCAR: An Attention-Aware 2-Bit KV Cache Quantization System for Long-Context LLM Serving
- 来源：MarkTechPost
- 发布时间：2026-05-25 21:24 UTC
- 链接：https://www.marktechpost.com/2026/05/25/together-ai-open-sources-oscar-an-attention-aware-2-bit-kv-cache-quantization-system-for-long-context-llm-serving/
摘要：Together AI has released OSCAR (Offline Spectral Covariance-Aware Rotation), an INT2 KV cache quantization method for long-context LLM serving. Unlike prior rotation-based approaches that apply data-oblivious Hadamard tr
2. Step by Step Guide to Build and Compare FedAvg and FedProx Federated Learning on Non-IID CIFAR-10 with NVIDIA FLARE
- 来源：MarkTechPost
- 发布时间：2026-05-25 20:26 UTC
- 链接：https://www.marktechpost.com/2026/05/25/step-by-step-guide-to-build-and-compare-fedavg-and-fedprox-federated-learning-on-non-iid-cifar-10-with-nvidia-flare/
摘要：In this tutorial, we build an advanced federated learning experiment with NVIDIA FLARE. We compare FedAvg and FedProx on a non-IID CIFAR-10 setup, where client data is split using a Dirichlet distribution to simulate rea
3. Pope Leo calls for being ‘profoundly human’ in the age of AI
- 来源：The Verge AI
- 发布时间：2026-05-25 15:05 UTC
- 链接：https://www.theverge.com/news/936945/pope-leo-letter-encyclical-ai-anthropic-labor-warfare
摘要：Pope Leo XIV warned of the risks of AI and unconstrained technological power in his first major papal document released on Monday. Magnifica Humanitas is the pope's manifesto on "safeguarding the human person in the time
4. Best Authentication Platforms for AI Agents and MCP Servers in 2026
- 来源：MarkTechPost
- 发布时间：2026-05-25 10:34 UTC
- 链接：https://www.marktechpost.com/2026/05/25/best-authentication-platforms-for-ai-agents-and-mcp-servers-in-2026/
摘要：As MCP crosses 97 million monthly SDK downloads and AI agents move into production workflows, authentication has become the most critical infrastructure decision teams face. This guide ranks the eight leading platforms —
5. WorkOS Releases auth.md: An Open Agent Registration Protocol Built on OAuth Standards
- 来源：MarkTechPost
- 发布时间：2026-05-25 07:38 UTC
- 链接：https://www.marktechpost.com/2026/05/25/workos-releases-auth-md-an-open-agent-registration-protocol-built-on-oauth-standards/
摘要：Most web applications still have no structured way for an AI agent to register. auth.md proposes a fix: a Markdown file apps publish at their domain that tells agents which registration flows are supported, which scopes
6. BOHM: Zero-Cost Hierarchical Attribution for Compound AI Systems
- 来源：arXiv cs.AI
- 发布时间：2026-05-25 04:00 UTC
- 链接：https://arxiv.org/abs/2605.22866
摘要：arXiv:2605.22866v1 Announce Type: new Abstract: Compound AI systems route tasks through hierarchies of specialised components. Attribution is dominated by Shapley-based methods (SHAP), which decompose a coalition value f
7. NeuroNL2LTL: A Neurosymbolic Framework for Natural Language Translation of Linear Temporal Logic
- 来源：arXiv cs.AI
- 发布时间：2026-05-25 04:00 UTC
- 链接：https://arxiv.org/abs/2605.22874
摘要：arXiv:2605.22874v1 Announce Type: new Abstract: Effectively translating between natural language (NL) and formal logics like Linear Temporal Logic (LTL) requires expertise that limits formal verification's reach in safet
8. RMA: an Agentic System for Research-Level Mathematical Problems
- 来源：arXiv cs.AI
- 发布时间：2026-05-25 04:00 UTC
- 链接：https://arxiv.org/abs/2605.22875
摘要：arXiv:2605.22875v1 Announce Type: new Abstract: We present $\textbf{Research Math Agents (RMA)}$, an agentic framework for automated reasoning on research-level mathematical problems. Unlike prior studies centered on com
9. SciAtlas: A Large-Scale Knowledge Graph for Automated Scientific Research
- 来源：arXiv cs.AI
- 发布时间：2026-05-25 04:00 UTC
- 链接：https://arxiv.org/abs/2605.22878
摘要：arXiv:2605.22878v1 Announce Type: new Abstract: The exponential growth of global academic output has confronted researchers and AI agents with an unprecedented ``information explosion,'' where fragmented and unstructured
10. Energy per Successful Goal: Goal-Level Energy Accounting for Agentic AI Systems
- 来源：arXiv cs.AI
- 发布时间：2026-05-25 04:00 UTC
- 链接：https://arxiv.org/abs/2605.22883
摘要：arXiv:2605.22883v1 Announce Type: new Abstract: Current AI energy benchmarks measure consumption at the granularity of a single model invocation or training run. For classical single-turn workloads this unit remains cohe
11. ImProver 2: Iteratively Self-Improving LMs for Neurosymbolic Proof Optimization
- 来源：arXiv cs.AI
- 发布时间：2026-05-25 04:00 UTC
- 链接：https://arxiv.org/abs/2605.22885
摘要：arXiv:2605.22885v1 Announce Type: new Abstract: Formal mathematics libraries are rapidly expanding, creating a growing need to refactor verified proofs for maintainability and to improve training data quality for neural
12. Mediative Fuzzy Logic: From Type-1 Foundations to Type-2, Type-3 and Quantum Extensions
- 来源：arXiv cs.AI
- 发布时间：2026-05-25 04:00 UTC
- 链接：https://arxiv.org/abs/2605.22900
摘要：arXiv:2605.22900v1 Announce Type: new Abstract: Mediative Fuzzy Logic was conceived as a practical scheme for reconciling hesitant or conflicting assessments in fuzzy control and decision-making. However, its logical and
13. EVE-Agent: Evidence-Verifiable Self-Evolving Agents
- 来源：arXiv cs.AI
- 发布时间：2026-05-25 04:00 UTC
- 链接：https://arxiv.org/abs/2605.22905
摘要：arXiv:2605.22905v1 Announce Type: new Abstract: Self-evolving agents should not train on examples they cannot justify. Data-free self-evolving search agents offer a scalable route to systems that generate their own quest
14. The Deterministic Horizon: Impossibility Results as Design Specifications for Trustworthy AI Systems
- 来源：arXiv cs.AI
- 发布时间：2026-05-25 04:00 UTC
- 链接：https://arxiv.org/abs/2605.23024
摘要：arXiv:2605.23024v1 Announce Type: new Abstract: Large language models now write software, draft legal documents, and produce clinical notes, yet fundamental limits, from Turing and Arrow to the No Free Lunch theorems, sh
15. PathCal: State-Aware Reflection-Marker Calibration for Efficient Reasoning
- 来源：arXiv cs.AI
- 发布时间：2026-05-25 04:00 UTC
- 链接：https://arxiv.org/abs/2605.23074
摘要：arXiv:2605.23074v1 Announce Type: new Abstract: The emergence of Large Reasoning Language Models (LRMs) has paved the way for tackling complex reasoning tasks through test-time scaling by generating long-form Chain-of-Th
16. Inductive Deductive Synthesis: Enabling AI to Generate Formally Verified Systems
- 来源：arXiv cs.AI
- 发布时间：2026-05-25 04:00 UTC
- 链接：https://arxiv.org/abs/2605.23109
摘要：arXiv:2605.23109v1 Announce Type: new Abstract: AI agents increasingly excel at generating, testing, and refining code. However, they fall short on tasks requiring formal guarantees of full coverage that testing alone ca
17. Redrawing the AI Map: A Theory of Accountability Boundaries in Agentic Ecosystems
- 来源：arXiv cs.AI
- 发布时间：2026-05-25 04:00 UTC
- 链接：https://arxiv.org/abs/2605.23179
摘要：arXiv:2605.23179v1 Announce Type: new Abstract: Agentic AI orchestrators reduce the interface and assembly costs of composing information systems capabilities across organizational boundaries, seemingly accelerating modu
18. AutoResearch AI: Towards AI-Powered Research Automation for Scientific Discovery
- 来源：arXiv cs.AI
- 发布时间：2026-05-25 04:00 UTC
- 链接：https://arxiv.org/abs/2605.23204
摘要：arXiv:2605.23204v1 Announce Type: new Abstract: Scientific research is being reshaped by AI systems that move beyond isolated assistance toward longer-horizon workflows spanning literature grounding, hypothesis generatio
19. Foundation Protocol: A Coordination Layer for Agentic Society
- 来源：arXiv cs.AI
- 发布时间：2026-05-25 04:00 UTC
- 链接：https://arxiv.org/abs/2605.23218
摘要：arXiv:2605.23218v1 Announce Type: new Abstract: Autonomous agents are moving from tools into a layer of social infrastructure: they browse, purchase, deploy software, manage systems, and increasingly interact with one an
20. GENSTRAT: Toward a Science of Strategic Reasoning in Large Language Models
- 来源：arXiv cs.AI
- 发布时间：2026-05-25 04:00 UTC
- 链接：https://arxiv.org/abs/2605.23238
摘要：arXiv:2605.23238v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly deployed as economic agents in marketplaces, auctions, and bidding settings. Anticipating their behavior in any specific deplo