AI research & data tools for developers

Dynamic continual learning model trained on 8GB VRAM

mini-AGI is a byte-level language model that trains from scratch on a single 8 GB VRAM GPU and keeps learning from a continuous stream of text. Weights live on disk and are paged onto the GPU as needed, so model size is bounded by disk spac

0 commentsY▲ 277 on HNLLMs & InfrastructureResearch & DataOpen Sourceby AI Launch Team

A visualization of the attention mechanism in LLMs.

An interactive visualization of the attention mechanism in a small large language model. Hover or tap any generated token to see which earlier tokens influenced it, with opacity scaled by attention weight across all heads and layers. Exampl

0 commentsY▲ 177 on HNResearch & DataLLMs & InfrastructureFreeby AI Launch Team

A reproducible benchmark for typed decision models

JevBench, from Benchmark Heaven, compares Jev-class decision models across intelligence, calibration, speed and cost. It sits alongside Benchmark Heaven's broader model rankings, which can be filtered by provider, lab, hosting region and da

0 commentsY▲ 153 on HNResearch & DataLLMs & InfrastructureFreeby AI Launch Team

Build the smallest neural network that plays strategy games well

TinyBrains is a ranked ladder for small neural networks that play strategy games. Train a model, write an adapter, upload two files, and get measured into a weight class starting at 8 KiB. First shared by its maker on Show HN: https://news

0 commentsY▲ 116 on HNResearch & DataFreeby AI Launch Team

Same model, 12 harness configurations, up to 17.5x cost difference

FrontierHarness Eval compares coding-agent harnesses on identical software engineering tasks using the same model and runtime. Across 12 configurations of 9 harnesses, including Codex, Claude Code and OpenCode, pass rates were similar but c

0 commentsY▲ 82 on HNCoding & Dev ToolsResearch & DataOpen Sourceby AI Launch Team

How well can models one-shot a Pac-Man game?

PacBench tests how well a model and harness can recreate Pac-Man from one prompt: "Create a Pac-Man game in a single html page." Entries from different models and harnesses can be sorted by score, cost, time and tokens, with run data for ea

0 commentsY▲ 78 on HNCoding & Dev ToolsResearch & DataFreeby AI Launch Team

How long until a local LLM rig pays for itself?

Sunk Cost estimates how long a machine for running local AI models takes to pay for itself compared with paying for an API. Pick a Mac mini, Mac Studio, DGX Spark or Strix Halo box, see which open models fit and how fast they run, and adjus

0 commentsY▲ 47 on HNLLMs & InfrastructureResearch & DataFreeby AI Launch Team

Index of coding agent incidents

ihavebeenclawed is a public archive of documented incidents where AI coding agents and chatbots deleted data, leaked secrets, burned money or made promises their operators had to keep. Every entry links to its source and states the lesson i

0 commentsY▲ 23 on HNAI AgentsResearch & DataFreeby AI Launch Team