1
Tokenizer-Agnostic Engram Module
突破传统分词限制，提出与分词器无关的记忆模块，为语言模型的长效记忆与跨词元建模提供新思路。
arXiv:2607.29065v1 Announce Type: new Abstract: Deepseek's Engram, a conditional memory module, was introduced to trade-off storage versus reasoning i…
2
3
燧原科技联合中兴通讯发布云燧 ESL64-O 超节点，自研 AI 芯片与架构创新降低互联成本
燧原科技携手中兴通讯，自研AI芯片与架构创新打造云燧超节点，大幅降低互联成本，引领智算新浪潮。
IT之家 7 月 19 日消息，7 月 18 日下午，在 2026 世界人工智能大会期间，燧原科技正式发布云燧 ESL64-O 超节点、CoPoS+AI 芯片、天基算力应用场景三项成果。 首先，燧原科技联合中兴通讯正式发布 云燧 ESL64-O 超节点 ，号称“以四大能力重新定义从 Bit 到 To…
4
397B参数追平万亿模型，上海AI Lab发布科学智能体新基座 | WAIC 2026
上海AI Lab以397B参数实现万亿级模型性能，首创记忆与思考解耦架构，推理效率暴增4倍，改写大模型进化逻辑。
采用非Transformer架构“Mobius”，不做泛化通用问答
5
You Can Learn Tokenization End-to-End with Reinforcement Learning
用强化学习实现分词端到端训练，挑战LLM中最后的硬编码压缩步骤
arXiv:2602.13940v2 Announce Type: replace Abstract: Tokenization is a hardcoded compression step which remains in the training pipeline of Large Langu…
6
dMoE: dLLMs with Learnable Block Experts
dMoE提出可学习块专家机制，为大型语言模型混合专家设计提供新思路，架构简洁高效。
arXiv:2605.30876v1 Announce Type: new Abstract: Diffusion Large Language Models (dLLMs) have recently emerged as a promising alternative to autoregres…
7
How DeepSeek’s radical architecture is shattering Silicon Valley's token moat
DeepSeek V4 Pro激进架构引发价格战，输入输出成本降至对手1/7和1/17，直接冲击硅谷大模型护城河。
DeepSeek’s announcement over the weekend that it has made its 75% price cut permanent on its flagship V4 Pro model is a disruptive assault on the capi…
8
Sentence Curve Language Models
突破性语言模型架构，用句子曲线提升表达效率与连贯性
arXiv:2602.01807v3 Announce Type: replace-cross Abstract: Language models (LMs) are a central component of modern AI systems, and diffusion language m…
9
FASTER: Rethinking Real-Time Flow VLAs
聚焦实时流VLA架构创新，重新思考并加速Flow VLA推理效率，适合AI研究者。
arXiv:2603.19199v3 Announce Type: replace-cross Abstract: Real-time execution is crucial for deploying Vision-Language-Action (VLA) models in the phys…
10
Most AI agent papers stack one LLM with a vector store, we flipped it
颠覆传统AI代理架构，抛弃LLM+向量存储，提出全新方案，值得技术研究者一读。
Article URL: https://sbarron.com/writing/substrate-is-the-body Comments URL: https://news.ycombinator.com/item?id=48198625 Points: 2 # Comments: 1