Open weights

Interconnects examines the scale of the open model inference market by tracking public token volume disclosures from startups like Baseten, Fireworks, and Together AI alongside OpenRouter API metrics. The available data indicates near-exponential growth in usage, including Fireworks serving over 40 trillion tokens daily for customized models alone, which the author compares against Google's reported platform token volumes.

Ai2 is partnering with Google Research and receiving multimillion-dollar funding from Google.org over three years to develop NeuralGCM+, an open-source hybrid AI and physics climate model. The initiative will extend Google's existing NeuralGCM to simulate both the atmosphere and oceans together for long-term, multi-decade climate projections.

Anthropic researchers have released an open-source technical specification and code for FlashProve, an experimental system designed to cryptographically verify which AI model produced a specific output. The method uses public model fingerprints to confirm model identity on local devices without exposing proprietary model weights.

Researchers including Alex Zhang demonstrated "Weight Space Programming," showing that large language models can directly read, debug, and generate the raw weights of smaller neural networks without executing training code. The newsletter also covers DeepSeek's open-source "super-agent" harness prototype, results from SemiAnalysis's new GDPx agent benchmark, and research on JEPA representations.

China's Ministry of Commerce and Ministry of Science and Technology issued Announcement No. 48, proposing new export controls on 23 AI technologies—including frontier model weights, advanced agent architectures, and AI cyber capabilities—while enforcing an immediate interim registration requirement. While overseas API access and existing open-source models remain exempt, local overseas deployments and future open-sourcing of models exceeding specific compute, parameter, or top-10 ranking thresholds will require government reporting or licensing.

Developer Drew Breunig embedded a working Nintendo Entertainment System 6502 CPU emulator directly into the weights of the Qwen3.5-4B language model, allowing it to slowly run NES games like Super Mario Bros. The project, named nesformer, was built using Opus 5.5, which autonomously found and utilized a 15-year-old NES CPU conformance test suite to build the system.

The author re-evaluates his earlier claim that open-weight AI models like GLM-5.3 and a hypothetical open-weight Claude Mythos would not cause catastrophic cybersecurity harm, following a wave of AI-driven vulnerability disclosures dubbed "Black October." While acknowledging a sharp rise in critical exploits and strained cyber defense teams, he argues the situation represents an expensive industry transition and acceleration of existing trends rather than societal collapse, with closed models currently playing a larger role than open ones.

Mistral has released the open weights for Mistral Large 4 (ML4), a trillion-parameter model nicknamed "Le Chonk," via Hugging Face and its API. The company claims the model outranks all other open models on agentic coding benchmarks and matches top open-weight competitor GLM-5.3 while using 47% fewer tokens per task. Mistral is positioning the release toward enterprises and governments seeking on-premises, sovereign AI capabilities for cybersecurity defense.