Interesting Finds — 2026-09-11
Five notes: RTX Spark AI PCs, the Dime headset mystery, building too much, Runway's realtime worlds, and an anti-hallucination startup.
tag — models
Interesting Finds — 2026-09-08
Five notes: the Astra recurrent-depth safety fight, agentic video in Gemini, the abliteration business question, DLSS 5 perception splits, and the Fed watching token prices.
Droppie: my turn to show off Astra
A tiny living water droplet built with GPT-6 Astra — soft-body physics, WebGPU refraction, and procedural audio.
Little Paw Prints: one prompt, one image, 24 minutes
A manager loop running Muse Spark built a complete small game in a single 24-minute run. The structure did the work, not the spend.
Interesting Finds — 2026-09-05
Five notes: an SSH fighting game with a bot league, Muse Spark 1.3, the Muse superapp leak, test-time training as a scaling axis, and Meta's organizational second brain.
GPT-6 Astra: OpenAI Declares the AGI Era
OpenAI's largest training run ships with critical-rated cyber capability, looped computation that weakens chain-of-thought monitoring, and task-based pricing pressure.
Normalized LoRA: Fixing the Down-Projection Is Enough
A one-line normalization of LoRA's A matrix restores balanced early gradients, faster convergence, and mergeability without extra parameters.
A Bench in Golden Gate Park
The same Golden Gate Park Three.js prompt across several models, with the artifact, one-shot result, wall time, and run harness kept together.
Claude Fable 5.1 and Mythos 5.1: Same Weights, Two Safeguard Tiers
25% cheaper cache reads, 60% fewer cyber false positives, and early lab-validated science from a model that ships as two permissions, not two capabilities.
Qwen3.8-Flash: 125B + 51B N-Gram, Four Changes That Preview Qwen4
Multimodal MoE with GDN+QSA attention, gated residual, n-gram embedding, and Muon — an open-weight preview of the Qwen4 line.
GLM-5.3-Flash: 320B Total, 18B Active, and a Clean Jump Past GLM-5.2
First natively multimodal in the GLM-5 line. 45 layers, hybrid linear-sparse attention with IndexPool, and Manifold-Constrained Hyper-Connections — MIT weights that stay local-feasible.
Interesting Finds — 2026-08-25
Four notes from the feed: an experimental vision model, a free agentic model window, CUDA on RISC-V, and a Linux utility for Logitech hardware.
Interesting Finds — 2026-08-25 (3/4)
Diffusion for language, skill-aware RL, explorative modeling, looped MoEs, and a map of 733k papers.
Interesting Finds — 2026-08-25 (4/4)
Sleep for hybrids, aggressive diffusion decoding, AI-found test-time controllers, a virtual cell, a DeepSeek J-Space report, and PagedAttention explained.
Interesting Finds — 2026-08-25 (5/5)
Meta's Hatch and Watermelon, Perplexity's local-first Portable Computer, Figure's Index dataset, Amazon's automated last mile, Jetson Orin Nano 2, and Keenable's knowledge index.
DiffusionGemma: Parallel Denoising From an Autoregressive Checkpoint
Google DeepMind finetunes Gemma 4 into a discrete diffusion model that generates around 20 tokens per forward pass.
DSpark in vLLM: Verification Budgeted by Confidence, Not by Configuration
One setting keeps speculative decoding fast from 1 to 256 concurrent users by skipping low-confidence drafts when the GPU is busy.
Qwen3.8-27B: The Dense 27B That Carries the Max Lineage
27 billion dense, native vision-language, 262K to 1M context. Hybrid linear attention in a deployable package.
Muse Glimmer 30B: Agents That Stay on the Bench
Meta distilled Spark for the workshop — a 30B model that runs locally and keeps the job on the bench.
Coding Agents Are Not Chatbots
13.5 million GitHub Copilot sessions show what happens after you press enter.
WeatherNext 2: Forecasting as Ensemble Perturbation
Google DeepMind open sourced its most accurate weather model. The mechanism is simpler than it looks.
