RightNow Research Lab
让模型与硬件的协同设计规模化

NVIDIA Inception 成员, arXiv 上的 8 篇 papers

推荐语
Director of Accelerated Computing, NVIDIA
研究

2607.20792cs.LG
Memoir: Should a Model Write to Its Memory While It Thinks?


2606.09682cs.LG
AutoMegaKernel: A Statically-Checked Agent Harness for Self-Retargeting Megakernel Synthesis
Jaber Jaber*, Osama Jaber, 2026

2604.02051cs.LG
Ouroboros: Dynamic Weight Generation for Recursive Transformers via Input-Conditioned LoRA Modulation
Jaber Jaber*, Osama Jaber, 2026

2603.29090cs.LG
HCLSM: Hierarchical Causal Latent State Machines for Object-Centric World Modeling
Jaber Jaber*, Osama Jaber, 2026

2603.21365cs.LG
TIDE: Token-Informed Depth Execution for Per-Token Early Exit in LLM Inference
Jaber Jaber*, Osama Jaber, 2026
Agent OS
OpenFang
Low-level agent execution with direct OS and GPU access.
RightNow-AI/openfang打开代码库
OpenFang: GitHub star 数
18,209
GitHub API 计数;图表为示意
Edge inference
PicoLM
Small inference runtime for memory-constrained edge devices.
RightNow-AI/picolm打开代码库
PicoLM: GitHub star 数
2,193
GitHub API 计数;图表为示意
Kernel search
AutoKernel
Profiles PyTorch models, tests Triton variants, and follows measured bottlenecks.
RightNow-AI/autokernel打开代码库
AutoKernel: GitHub star 数
1,564
GitHub API 计数;图表为示意



