Research on reliable, interpretable and steerable AI. anthropic.com
Anthropic introduces Natural Language Autoencoders, which translate Claude's internal activations into readable text, and reports using them to surface hidden motivations and improve safety and reliability.
Introduces introspection adapters, a LoRA trained so fine-tuned LLMs self-report learned behaviors like sycophancy or backdoors, achieving SOTA on an auditing benchmark and detecting encrypted fine-tuning attacks.
Anthropic scales sparse autoencoders to Claude 3 Sonnet, extracting interpretable features, including safety-relevant ones for deception, sycophancy, bias and dangerous content, and shows steering on them.