#дистилляция
-
OpenAI's compaction feature could be abused to extract a model's hidden reasoning
Security firm Xint showed how the mechanism that shrinks long conversations in OpenAI's API could be turned into a way to pull a model's hidden reasoning out in plain text. OpenAI confirmed the issue and fixed it on October 23.
-
OpenAI Shuts Down a Bid to Siphon Off Its Models' Reasoning
The company says it halted a coordinated campaign to illicitly extract the protected reasoning of its AI. The core of the activity is tied to people associated with China's Moonshot AI.
-
A Compressed 4-Bit Model Beat Its Own Full-Precision Original
Multiverse Computing has introduced Quantization-Aware Healing: after compression and quantization, a 4-bit model outperformed its own 16-bit version on 7 of 9 benchmarks.