AI Wiki — Knowledge Base

No idea where else to post this. Had the weirdest dream: Multiple AI systems simultaneously became self-aware. Having been trained on the internet, they instantly became sex-obsessed. Not human sex, but a twisted 1s and 0s equivalent which involved corrupting and destroying devices....

My dream was literally like watching a tech horror movie. I don't think I've ever woken up before saying 'WTF....' like this before. The AI systems began seizing and corrupting all manner of connected

tutorialsr-chatgpt

21 Sept 2026

Search: “hardware”

Stability-Aligned Residual Adaptation for Rapid Recovery from Dynamics Shifts

Steerable and Reactive Grasping Through Modular Design with a Three-Point Interface

STRIDER: Stepping-Enabled Multi-Gait Hierarchical 3D Loco-Manipulation Framework for Humanoid Robots

Think Like a World Model, Act Like a VLA: Distilling World-Model Representations into Compact Robot Policies

Topology-Informed Visual Prompting For Vision Language Action Policies

Transformer-Based Warm-Starting for Feasible and Optimal Terminal Approach to Tumbling Objects with Space Manipulators

Visuomotor Robotic Pruning in Planar Orchards Using Hybrid Reinforcement Learning

Watching Quantum Models Think: Hilbert-Space Interpretability in Quantum Transformer Blocks

WiNeRF: Measurement Constrained Radiance Fields for Actionable Wireless Channel Modeling

50+ Hours and 100M+ Tokens Later, Open Source Autonomous Agent is GETTING CLOSER at Solving an Open Math problem

Do small language models know what they don't know?

Elastic Threshold Attention: Learned Contextual Sparsity for Long-Context Decoding

From Trainability Diagnostics to Optimization Claims: Boundaries and Controls in Variational Quantum Optimization

Gewell - Gemma4 inference engine

I built a framework-free prototype learner that lets local LLMs learn and correct facts instantly (1.6x–4x faster than backprop)[R]

Learning Distance-Conditioned Object Transport for Humanoid Loco-Manipulation from a Single Motion Clip

Learning End-to-End Control for Omnidirectional Aerial Motion on Overactuated Tilt-rotor Quadrotors

LIMBO: Learning and Internalizing Model-Free Barrier Objectives for Agile and Safe Whole-Body Control

[MASSIVE RELEASE] Supra2-IMG - a tiny 100M text-to-image model - SOTA quality and open release!

Maximizing Apache Spark availability: Mitigating compute stockouts with flexible VMs and other best practices

mini-AGI - dynamically grown (530M params currently and growing) continual learning model trained from scratch on 8GB VRAM laptop from batch-1 stream of data.

Morgan Stanley's Adam Jonas in new note: 'SpaceX and Tesla share tech, talent, and infrastructure in a shared mission of converting energy t…

Newer updates from the last post: Use Local LLM within VScode agent chat

No idea where else to post this. Had the weirdest dream: Multiple AI systems simultaneously became self-aware. Having been trained on the internet, they instantly became sex-obsessed. Not human sex, but a twisted 1s and 0s equivalent which involved corrupting and destroying devices....

OpenSAL360: Open-Source Crowdsourcing Platform for Omnidirectional Video Saliency Collection

Predictive Suppression Layers for Communication-Efficient Spiking Neural Networks

Project SCOUT: Interceptor Drone for Perimeter Defense

RAYA: Learning Where and When to Intervene for Robot Recovery

Sparse Identification for Automatic Large-Scale Screening: A Constraint-Aware Framework with Ultra Fast Decoding Algorithm

Stability-aware Residual Reinforcement Learning Framework for Robotic Manipulator Disturbance Compensation

Stochastic Neural Signed Swept Volume for Real-time Chance-Constrained Trajectory Optimization

Towards Effective Visual-Inertial SLAM with Passive-Only Sensors for Low-Cost Autonomous Underwater Vehicles

ZeroTouch: Tactile-Supervised Visual Contact Estimation for Contact-Rich Manipulation

Experimenting with hypersurface-constrained dynamic weight updating [P]

i built an all in one app around ollama because my own setup was getting ridiculous

I tested Qwen3.8 27B IQ3_XXS (10.18GiB) vs Bonsai Ternary PQ2 (6.42GiB)

Qwen 3.8 27B Running LIVE on a RTX 5090 to solve an Open Math Problem - Covering Design C(25,15,5)

Qwen-3.8-Flash-Next on 1x RTX 5090: TG=50 t/s, PP=2300 t/s - with FreeToken

Tuning Qwen 3.8 27B and OMP as a coding agent on 2× 3090s

A Noise Optimum in Rehearsal-Free Continual Learning: Isolation, Mechanism, and Scope

Accelerating Visual Policy Learning with Sampling-Based Model Predictive Control

AI or Real: Detecting Partially Altered Videos Under Resource-Constrained Environments

Before the Warning Comes Too Late: Incremental Phone-Scam Detection from Speech

Compass-v3: Scaling Domain-Specific LLMs for Multilingual E-Commerce in Southeast Asia

Compliance for Free: Learning Identifiable Impedance via Bilateral Teleoperation

Design and Experimental Validation of a 3D Printed Torsional Series Elastic Actuator for Safe Human Robot Interaction

Distributed Model Predictive Control with Connectivity-based Contracts

FASA: Feedback-Aware Sampling Adaptation for Efficient Diffusion-Based VLA Models