#llm

🔤 Wordmaze: a one-shot, verifier-graded puzzle for RL Jun 3

🔍 Agents Can Reason. They Still Can't Really Search. Mar 17

📏 Bits per byte (BPB) for LLMs: tokenizer-agnostic loss Oct 15

🔀 GPT-5 router: why chat apps will route across models Aug 13

☁️ AWS Bedrock Converse API: one endpoint for Claude, Llama, and Mistral Jun 13