lemmy VS openai-cookbook

I'm aware of how it works since I have been spending a lot of time over the last two months working with Claude's internals. If you have spent some time with it, you know that it is a transpiled and minified mess that is annoyingly hard to detangle. I'm very happy that claude-trace (and claude-bridge [1]) exists because it makes it much easier to work with the internals of Claude than if you have to decompile it yourself.

[1]: https://github.com/badlogic/lemmy/tree/main/apps/claude-brid...

Also check out claude-trace, which injects fetch hooks to get at the data: https://github.com/badlogic/lemmy/tree/main/apps/claude-trac...

======

https://github.com/badlogic/lemmy/blob/main/packages/lemmy/t...

Obviously a toy example, but never the less interesting. Anthropic's models provide thinking tokens, so we can get a fuzzy glimpse into its reasoning. Sometimes the models understand they are role playing, sometimes they don't. Even if they do say they know they are role playing, they often eventually do something malicious or at least self-preserving.

OpenAIs reasoning models don't provide thinking tokens, but you can derive their reasoning from their actions. o3 and o1 will both do malicious or self-preserving things. Couldn't get o4-mini to do anything bad. Non-reasoning models like 4.1 and 4o also don't seem to have any self-preservation.

Since not a lot of info is supplied to the models, they all start hallucinating email addresses and people after a few turns. Role-players gonna role-play I guess.

I documented a bunch of conversations in this Bluesky thread.

OpenAI Cookbook Prompt engineering tips, context tricks, embeddings, and more from the people who built the models https://github.com/openai/openai-cookbook

The OpenAI Cookbook isn’t just an API reference it’s packed with real examples: prompt chaining, code debugging, SQL generation, code review flows, even CLI wrappers. Between this and their official docs, I noted dozens of structured developer use cases.

Anyone who has recently worked on embedding model finetuning, any useful tools you'd recommend (both for dataset curation and actual finetuning)? Any models you'd recommend as especially good for finetuning?

I'm interested in both full model finetunes, and downstream matrix optimization as done in [1].

[1] https://github.com/openai/openai-cookbook/blob/main/examples...

LLM comparison tracker

Sam Altman wrote in February that GPT-4.5 would be "our last non-chain-of-thought model" [1], but GPT-4.1 also does not have internal chain-of-thought [2].

It seems like OpenAI keeps changing its plans. Deprecating GPT-4.5 less than 2 months after introducing it also seems unlikely to be the original plan. Changing plans is necessarily a bad thing, but I wonder why.

Did they not expect this model to turn out as well as it did?

[1] https://x.com/sama/status/1889755723078443244

[2] https://github.com/openai/openai-cookbook/blob/6a47d53c967a0...

Resource: OpenAI Cookbook