True Positive Weekly #117

The most important artificial intelligence and machine learning news and articles

Hey, in this issue: are bigger AI models better stock pickers?; a foundation model to predict and capture human cognition; LSTM or Transformer as malware packer; on the differences between MCP and API; transformers for black-box optimization; accelerating Llama 4 MoE inference; an inference framework for 1-bit LLMs; and more.

• [FT] Are bigger AI models better stock pickers? Maybe, but probably not

• [Nature] A foundation model to predict and capture human cognition

• LSTM or Transformer as malware packer

• MCP vs API: On the differences between MCP and API

• You can just predict the optimum: Transformers for black-box optimization

• [Tool] Official inference framework for 1-bit LLMs from Microsoft

• Vector search in the real world: How to filter efficiently without killing recall

• MetaShuffling: Accelerating Llama 4 MoE inference

Enjoy the newsletter? Please help us make it bigger and better by sharing it with colleagues and friends.