AI News Daily

Researchers at the Technical University of Munich gave Meta's video world model the kind of test psychologists give infants. On rendered scenes, the part of the model that sees knows a ball is inside a closed box, while the part that fills in what it cannot see lets the ball go. A short training run fixes that, yet a standard benchmark score rises just as much after training that never teaches it.

The new classifier keeps Qwen3-Omni's image, audio and video inputs while returning typed answers instead of generated prose. Cloudflare also cut Clef-flash's price and context window.

A preprint finds that language models infer relationships from ambiguous evidence more readily than human participants. Its controlled tests connect the effect to frequency-sensitive internal features.

A preprint finds that models do not reliably report their own quantization from text behaviour. Internal probes can detect known methods, but fail to generalise to an unseen one.

Maurice Heumann says a three-month agent project began making reliable progress only after every function was compiled and compared with the original binary.

Sakana AI and UC Berkeley use one model to design, execute and judge multi-agent workflows, then train that model on the best team outputs.

The open-source runner keeps selected model experts in system memory and moves them to a consumer GPU as needed. Its one-click package exposes local OpenAI- and Anthropic-compatible APIs.

The open-weight decision model answers bounded questions in one forward pass and returns probabilities. Its strongest speed and accuracy figures come from H2O.ai's own tests.

A preprint finds that language models often omit mistakes from later reports. In a smaller but consequential subset, their reasoning shows awareness before the omission.

A preprint separates a reversible loss of access from the slower destruction of stored facts. The distinction could change how continual-learning failures are diagnosed.

MOLT Sinai routes software work through specialised agents, mandatory plans and adversarial review. The company reports seven reverts over five months.

The open-source assistant combines chat, editable memory, tasks and reminders inside one user's Cloudflare account. Its deployment is private, but not local.