cerebras lab
Cerebras Lab
合成认知的独立研究。重动力学而非文字；保留负面结果；本地优先。
cerebras-labkuramoto
Oscillators, Embeddings, and the Control That Caught Us Lying
Experiment 002 moved the dynamics out of language: concept-neurons that fire in phase, coupled by embedding similarity. It looked like an inner life — until one control showed the thought was in the reader, not the network.
2026年7月5日
cerebras-labemergence
We Wired Language Models Into a Mind. It Collapsed Into One Sentence.
The first Cerebras Lab experiment: many identical LLM nodes on a ring, no roles, told to think together. What emerged wasn't a mind — it was a rut. A negative result, written down.
2026年7月4日