Instituto de Ciencias del Mar y Limnología
PulseAugur coverage of Instituto de Ciencias del Mar y Limnología — every cluster mentioning Instituto de Ciencias del Mar y Limnología across labs, papers, and developer communities, ranked by signal.
1 天有情绪数据
-
Meitu launches seamless text editing, Xiaomi patents driving safety tech
Meitu has integrated its research on scene text editing into its Meitu Design Studio app and Meitu Xiuxiu PC version with a new 'seamless text modification' feature. This function supports multiple languages including C…
-
Meitu AI research accepted to top conferences, powers new editing features
Meitu's AI research arm, MT Lab, has had six papers accepted into major international conferences including ICLR, CVPR, and ICML. One paper on scene text editing, accepted by ICML 2026, has already been integrated into …
-
New LLM research tackles factuality with semantic clustering and conformal prediction
Researchers are exploring novel methods to combat Large Language Model (LLM) hallucinations and improve their factuality. Semantic Entropy analyzes answer variations to detect confabulations, while Linguistic Calibratio…
-
苹果研究人员提出缓存共享以降低 LLM 服务成本
Apple Machine Learning Research 发布了一篇论文，详细介绍了一种名为 Stochastic KV Routing 的新方法，以减小 transformer 语言模型的内存占用。该技术侧重于优化 KV 缓存的深度维度，而不是仅仅进行时间压缩或淘汰。通过训练层随机关注先前层的 KV 状态，模型能够适应各种缓存共享策略而不会丢失信息，有可能在显著降低内存使用量的同时保持或提高性能。
-
Apple advances normalizing flows, researchers explore denoising and state estimation
Apple Machine Learning Research has introduced iTARFlow, an advancement in Normalizing Flow generative models that maintains a likelihood-based objective and uses an iterative denoising procedure for sampling. This meth…
-
AI agents gain intelligence via metacognition and prompt optimization
Recent research explores advanced agent architectures that move beyond simple retry loops for complex tasks. Studies like "Supervising Ralph Wiggum" demonstrate that separating metacognitive critique into a distinct age…
-
Apple 的 SeedLM 使用伪随机生成器压缩大语言模型权重
研究人员开发了 SeedLM，一种用于大语言模型的新型训练后压缩技术，该技术利用伪随机生成器种子来编码模型权重。该方法旨在通过在推理过程中即时生成权重矩阵来降低大语言模型的高运行时成本，从而减少内存访问并提高内存密集型任务的速度。SeedLM 通过用计算换取更少的内存访问来实现这一点，并且显著的优点是不需要校准数据，在各种任务上都能很好地泛化，并且在显著的压缩水平下仍能保持与 FP16 基线相当的准确性。