Skip to content
View rmems's full-sized avatar
13 followers · 20 following

Achievements

Achievement: Pair Extraordinairex4Achievement: Pull Sharkx3Achievement: QuickdrawAchievement: YOLO

Achievements

Achievement: Pair Extraordinairex4Achievement: Pull Sharkx3Achievement: QuickdrawAchievement: YOLO

Highlights

  • Pro

Organizations

@Limen-Neural

Block or report rmems

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
.github/profile/README.md

Raul Montoya Cardenas

San Marcos, Texas · montoyaraul34@gmail.com

Making large models run on hardware you can actually own.

I research SAAQ (Spiking Adaptive Activity Quantization): ways to study and compress large MoE models with spiking / neuromorphic ideas, mostly on hardware I can run myself (e.g. an RTX 5080). I am still learning Julia, Rust, Python, and CUDA.

Repos: rmems · libraries: Limen Neural

Projects
magere-brug SAAQ lab (recipes, manifests, experiments)
corinth-canal SAAQ reference pipeline (Rust)
xai-dissect Inspect open-weight Grok checkpoints
grok-ozempic Grok-scale SNN-style quantization experiments
Surrogate_Viz.jl Symbolic regression on telemetry
XAIDissect_Viz.jl Viz for xai-dissect reports
gaming-telemetry / Theseus-Quarry Telemetry for neuromorphic work
operation-prometheus Engineering-trajectory datasets for local agents

More under rmems (training forge, benchmarks, experimental SNN-HFT research that is not live trading, worktree-hive early/experimental, etc.).

How I work

I normally go from issue → pull request, either from a local CLI or through Linear (cloud).

  • Agents I use day to day: Grok Build, Codex, Claude, Cursor, Devin, and others — plus shared memory (Ogham / Chroma). Humans merge.
  • PR review (rough order): Codex → CodeRabbit → Devin when I want a strong extra pass → Cursor Bugbot · Copilot · CodeAnt · Qodana (local + cloud).
  • CI: GitHub Actions on pull requests and pushes to main research repos.

Currently

README updated with Grok Build: Grok 4.5 (high)

Popular repositories Loading

  1. metabolic-ledger metabolic-ledger Public

    Bio-inspired simulation ledger: ATP cellular energy metaphors, adaptive Kelly energy commitment, and metabolic cost tracking for multi-asset SNN portfolios.

    Rust 1

  2. Ship-of-Theseus-HPC Ship-of-Theseus-HPC Public

    Localized HPC node for Bio-MEMS simulation, RTL design (SystemVerilog/Rust), and hardware diagnostics. Documentation for the 'Ship of Theseus' workstation.

  3. LiquidCortex.jl LiquidCortex.jl Public

    GPU-accelerated sparse Liquid State Machine for neuromorphic inference — 65k-neuron/lobe CUDA LSM with OU-SDE dynamics and STDP learning

    Julia

  4. NeuroPulse.jl NeuroPulse.jl Public

    NERO: Neuromorphic Evaluation of Relevance and Orchestration — multi-lobe SNN relevance scoring with cross-lobe inhibition and softmax normalisation

    Julia 1

  5. limbic-critic limbic-critic Public

    Modulator Mapping: into constrained f32 vectors representing Dopamine (reward), Serotonin (risk/patience), and Cortisol (stress/telemetry)

    Rust

  6. SpikeStream.jl SpikeStream.jl Public

    Streaming time-series feature extraction for spiking neural networks: Hurst exponent, Hawkes intensity, GBM surprise Z-score — SNN-compatible output ranges, zero allocation

    Julia