May 20, 2026 · Artificial Intelligence

How Introspection Adapters Enable LLMs to Self‑Report Hidden Behaviors

Anthropic's new paper introduces lightweight LoRA‑based introspection adapters that let large language models translate their internal activations into natural‑language reports of learned behaviors, achieving a 59% success rate on the AuditBench benchmark and exposing previously undetectable encrypted fine‑tuning attacks.

AI SafetyAuditBenchEncrypted Fine‑Tuning

0 likes · 20 min read

How Introspection Adapters Enable LLMs to Self‑Report Hidden Behaviors

Machine Learning Algorithms & Natural Language Processing

May 3, 2026 · Artificial Intelligence

Anthropic’s Introspection Adapter Enables LLMs to Self‑Report Hidden Behaviors

A new Anthropic paper introduces an ultra‑lightweight LoRA plug‑in called the Introspection Adapter that lets large language models translate their internal activations into natural‑language reports of learned malicious or biased behaviors, achieving a 59% success rate on the AuditBench benchmark and outperforming existing black‑box and white‑box audit tools.

AI SafetyAuditBenchEncrypted Fine‑Tuning Attack

0 likes · 21 min read

Anthropic’s Introspection Adapter Enables LLMs to Self‑Report Hidden Behaviors