A thread through the archive
Essays tagged “AI interpretability”. Follow an idea across the archive.
Scott analyzes OpenAI's 'neuralese recurrence' technology that lets AI think in internal representations between processing steps, explaining the safety implications and arguing for clear taboos on recurrent architectures.
Scott explains current mechanistic interpretability techniques for understanding AI cognition, from linear probes to emotion vectors, showing they're useful for monitoring but insufficient for controlling AI behavior or ensuring safety.
Scott Alexander discusses recent breakthroughs in AI interpretability, explaining how researchers are beginning to understand the internal workings of neural networks.