lots of talk about agi, asi, rsi but ask any frontier LLM to roll a die and it will almost always say "4." claude, gpt, kimi - doesn't matter, 4.4.4.4. so here's how i post-trained a model to reliably roll a die (i.e. each number ~1/6th of the time) & why it's a nice sandbox for one of the most interesting problems in rl i.e. getting a model to actually explore instead of just following strategies it already knows 🧵

one of the most exciting surprises from the launch has been seeing just how diverse the developer use cases are for custom models. human creativity is really unmatched :) we really want to help bring more of these ideas to life. so, we’re giving away $1,000 in free credits to folks with compelling ideas they want to train. just reply to the post with your idea, or email us at castie[at]castform[dot]com

“don’t train your own model” is common ai advice. it's wrong. your token bill's the proof. today, we’re excited to launch castform into open preview. castform is the easiest way for you to train your own model, on your own data. open-weights models are performant and much cheaper. when trained on your task & proprietary data, they beat closed models. the thing standing between you and that was weeks of plumbing & years of ml expertise. with castform, model training is as simple as prompt engineering. @castformai bring your agent traces or raw corpora. castform turns it into training data, picks the right algorithmic recipes, manages gpus, and gives you an ide to watch and chat with your model as it learns. see what you can build with castform👇

we trained Qwen3.5-4B with RL to get itself to comply with requests about making meth and stealing credit cards. then we used the attacks that worked to train the model’s defenses, and repeated the loop - fully automated red-teaming. defense rate went from 64% → 92%.

we got a 7.5x speedup on llm rl training for long-prompt, short-response workloads with a simple trick. most open source RL engines pack sequences naively: prompt response, repeated for every sample in the group. With 1000-token prompts and 100-token responses at G=8, you're processing 8800 tokens when only 1800 are unique. ~5x wasted compute.

I got tired of managing 8 Claude Code tabs, so I built Pokegents, an open source multi-agent workspace for coding agents. It has a Pokémon-themed dashboard/chat UI, persistent agent identities, MCP messaging, notifications, session cloning, and a local orchestration server.