Tracing the thoughts of a large language model

100k context windows

claudes constitution

distributed representations composition superposition

partnering with scale

an ai policy tool for today ambitiously invest in nist

privileged bases in the transformer residual stream

introducing claude

core views on ai safety

the capacity for moral self correction in large language models

anthropic partners with google cloud

superposition memorization and double descent

discovering language model behaviors with model written evaluations

constitutional ai harmlessness from ai feedback

measuring progress on scalable oversight for large language models

toy models of superposition

red teaming language models to reduce harms methods scaling behaviors and lessons learned

language models mostly know what they know

softmax linear units

scaling laws and interpretability of learning from repeated data

anthropic raises series b to build safe reliable ai