AI Paper Summary

Salesforce AI research team present FOFPred, a language driven future optical flow prediction framework that connects large vision language models with diffusion transformers for dense motion forecasting in control and video generation settings. FOFPred takes one or more images and a natural language instruction such as ‘moving the bottle from right to left’ and predicts…

As context lengths move into tens and hundreds of thousands of tokens, the key value cache in transformer decoders becomes a primary deployment bottleneck. The cache stores keys and values for every layer and head with shape (2, L, H, T, D). For a vanilla transformer such as Llama1-65B, the cache reaches about 335 GB…

A team of Stanford Medicine researchers have introduced SleepFM Clinical, a multimodal sleep foundation model that learns from clinical polysomnography and predicts long term disease risk from a single night of sleep. The research work is published in Nature Medicine and the team has released the clinical code as the open source sleepfm-clinical repository on…

Tencent Hunyuan researchers have released HY-MT1.5, a multilingual machine translation family that targets both mobile devices and cloud systems with the same training recipe and metrics. HY-MT1.5 consists of 2 translation models, HY-MT1.5-1.8B and HY-MT1.5-7B, supports mutual translation across 33 languages with 5 ethnic and dialect variations, and is available on GitHub and Hugging Face…

Can a 3B model deliver 30B class reasoning by fixing the training recipe instead of scaling parameters? Nanbeige LLM Lab at Boss Zhipin has released Nanbeige4-3B, a 3B parameter small language model family trained with an unusually heavy emphasis on data quality, curriculum scheduling, distillation, and reinforcement learning. The research team ships 2 primary checkpoints,…