🔄 In a Training Loop

SSH PRO

ProCreations

·

https://ssh.codes

AI & ML interests

Making AI’s small, accessible, and easy to run on consumer hardware

Recent Activity

posted an update about 13 hours ago

check out this cool webgpu thing i made! https://huggingface.co/spaces/ProCreations/maple-webgpu runs this model https://huggingface.co/deepgrove/maple-preview with cool performance i found today on WebGPU locally on your computer with very fast speeds. made with a mix of gpt 5.6 sol + opus 5 ultracode, its amazing what agents can do now, just a year ago this wouldve been impossible with them.

liked a model about 13 hours ago

deepgrove/maple-preview

liked a model about 14 hours ago

ProCreations/maple-preview-webgpu-v2

View all activity

Organizations

ProCreations 's Spaces 353

Reproduction: Prototype-Grounded Concept Models

Native PGCM prototype and intervention reproduction

Reproduction - On the Convergence Rate of LoRA Gradient Descent

Explore research logbook and sync with coding agents

Reproduction - Theory of Continual Learning with Gradient Descent

Explore research logbook and sync with an AI collaborator

Reproduction - Benign Overfitting in Adversarial ViTs

Exact ViT theorem, softmax and released-ablation audit

Reproduction - FormulaCode Agentic Optimization on Large Codebases

Full released-data audit of six FormulaCode claims

Reproduction: Bandit Social Learning Dynamics with Exploration Episodes

Native EpiBSL exploration-failure reproduction

Reproduction: Tighter GP-Bandit Regret Bounds on Hyperspheres

Native GP-bandit hypersphere audit

Reproduction - Doubly Regularized Robust MDPs

Exact theorem, ridge, UCB and robust-control audit

Reproduction: Human-AI Collaborative Uncertainty Quantification

Exact collaborative uncertainty-quantification audit

Reproduction: Full-Batch GD versus One-Pass SGD

Direct full-batch versus one-pass sample-separation audit

Reproduction - Delayed Momentum Aggregation

Exact DeMoA source, formula and Byzantine-majority audit

Reproduction: Prototype-Driven Bayesian Crowd Aggregation

Native PTBCC crowd-aggregation audit

Repro - CAffNet Hard Constraints

Explore and sync project logs with an AI coding agent

Adaptive Sensing for Principal Eigenvectors

View judge‑facing evidence scorecard

Tight Stability Bounds for Robust Distributed Learning

Explore reproducible robustness experiments for distributed learning

Reproduction - Voronoi Probabilistic Circuits

Voronoi probabilistic-circuit audit

Reproduction: Universal Representation of Generalized Convex Functions and Their Gradients

Native generalized-convex audit

Reproduction: Gaussian Universality Breakdown in High-Dimensional ERM

Direct Gaussian-universality ERM audit

Falsification logbook — Evaluating Relational Reasoning in LLMs

Exact source-data falsification of all six REL claims

Reproduction: Newton Step and Influence Function Accuracy

Exact logistic attribution audit

Reproduction: All ERMs Can Fail in Linear Dimension

Direct ERM and GD lower-bound reproduction

Reproduction Diffusion Model Collapse

Explore and collaborate on a research logbook

Reproduction: Committed Q-learning for Reactive RL

Direct native Committed Q-learning reproduction

Reproduction Efficient Preference Poisoning

Explore experiment logs and collaborate with an AI agent