CONNECTED RESEARCH

Knowledge map

Explore research through connected companies, technologies and questions. Every relationship links to its evidence.

55 / 55 articles with reviewed reader guides · 202 reviewed relationships

GenEval2 execution success rate

GenEval2 execution success rateGenEval2 execution success rDesigner-RSIDesigner-RSI

Select a connected node to continue exploring. Open the evidence below to distinguish reporting, analysis and hypotheses.

Reported evidence

Designer-RSI → changes → GenEval2 execution success rate

The abstract reports a rise in GenEval2 execution success from 72.7% to 99.3% on Claude Sonnet 4 under the complete Designer-RSI setup. This is an experimental result, not an isolated causal estimate for memory expansion, revision, or replay checks, and not a production acceptance rate.

Read the evidence
The abstract reports that, over five rounds involving 1,406 real user briefs and 1,869 automatically graded execution trajectories, the system increased its skill count from 76 to 139 without weight updates or human labels. GenEval2 execution success on Claude-Sonnet-4 rose from 72.7% to 99.3%, and generation quality increased by 11.99 points.[S4]

[S4] arXiv

Read the full research

Connected research

ai

From Cloud Access to Agent Memory: Delivery Paths and Conditions for Reusing AI Models

What do cloud model access, enterprise Agent service entry points, tool-call formats, and external skill memory each contribute—and what must still be tested before reuse?

The article separates access and configuration announcements, service listings, interface rules, and experimental results rather than treating them as one model ranking. Its most detailed mechanism example is Designer-RSI’s frozen-weight Agent with reusable skills and replay-gated memory updates.