Media report · AI-ASSISTED BRIEF
Anthropic Research Details Agent Memory Architecture and 'Dreaming' Consolidation for Claude
Report publication · YYSI publication
According to IT Home, Anthropic published research on agent memory, describing a four-layer design of session logs, memory notes, knowledge representations and injected context, plus a 'dreaming' mechanism where Claude reviews logs between sessions to correct errors and distill rules. On the CL-Bench database QA task, memory-enabled Claude Opus 5.5 and Sonnet 5.5 scored 0.838 and 0.833 versus 0.655 and 0.644 without memory, per the company.
Source & context
www.ithome.com · Read the original source ↗
This AI-assisted brief attributes information to the linked media report. It is not an official announcement or independent YYSI reporting. Consult the original report and its evidence.