9 papers
Memoir: Should a Model Write to Its Memory While It Thinks?
Jaber Jaber, Osama Jaber
Memoir combines per-sample fast memory, shared slow parameters, variable-depth latent recurrence, and a future-latent energy objective. We test its riskiest coupling: each ponderin…
Auto: The AGI Compiler
Jaber Jaber, Osama Jaber
Every LLM agent run re-derives its behavior token by token on a frontier model: brilliant, expensive, slow, and unbounded. We present Auto, a compiler that records live agent behav…
AutoMegaKernel: A Statically-Checked Agent Harness for Self-Retargeting Megakernel Synthesis
Jaber Jaber, Osama Jaber
AutoMegaKernel (AMK) compiles a HuggingFace Llama-family model into a single persistent cooperative CUDA kernel that runs the whole forward pass in one launch, with no per-model ha…
StreamIndex: Memory-Bounded Compressed Sparse Attention via Streaming Top-k
Jaber Jaber, Osama Jaber
DeepSeek-V3.2 and V4 introduce Compressed Sparse Attention (CSA): a lightning indexer (a learned scoring projection over compressed keys) scores them, the top-k are selected per qu…
RightNow-Arabic-0.5B-Turbo: An Open Sub-1B Arabic Language Model via Vocabulary Injection and Edge-First Deployment
Jaber Jaber, Osama Jaber
Open Arabic large language models split into two classes: sub-1B multilingual models that treat Arabic as an afterthought (Qwen2.5-0.5B, Falcon-H1-0.5B), and 7B-70B Arabic-speciali…
Ouroboros: Dynamic Weight Generation for Recursive Transformers via Input-Conditioned LoRA Modulation
Jaber Jaber, Osama Jaber
Recursive transformers reuse a shared weight block across multiple depth steps, trading parameters for compute. A core limitation: every step applies the same transformation, preve…
HCLSM: Hierarchical Causal Latent State Machines for Object-Centric World Modeling
Jaber Jaber, Osama Jaber
World models that predict future states from video remain limited by flat latent representations that entangle objects, ignore causal structure, and collapse temporal dynamics into…
TIDE: Token-Informed Depth Execution for Per-Token Early Exit in LLM Inference
Jaber Jaber, Osama Jaber
Large language models run every token through every layer, regardless of difficulty. We present TIDE, a post-training system that attaches tiny learned routers at periodic checkpoi…
AutoKernel: Autonomous GPU Kernel Optimization via Iterative Agent-Driven Search
Jaber Jaber, Osama Jaber
Writing high-performance GPU kernels is among the most labor-intensive tasks in machine learning systems engineering. We present AutoKernel, an open-source framework that applies a…