Today for AI
HOT RADAR
agentHEAT 13.3°

AI CLUSTERED EVENT · 10/12/2026

AWS and HSBC Introduce Double Ratchet: Self-Evolving Graders Enable Unsupervised Agent Improvement from 10 Samples

1 reports archived1 independent sourcesupdated 10/12/2026, 14:11:41
Synthesis & Latest Updates
1 Sources Cross-Validated

A joint paper by AWS and HSBC introduces the 'Double Ratchet' framework, eliminating the need for human-written graders in self-improving agents. By co-evolving the grader with agent skills using only 10 labeled examples to build rule-based checks, the method achieves training performance comparable to ground truth or rubrics (retaining 88-110% of gains). This significantly lowers the barrier for unsupervised agent self-improvement.

LATEST/A joint paper by AWS and HSBC introduces the 'Double Ratchet' framework, eliminating the need for human-written graders in self-improving agents. By co-evolving the grader with agent skills using only 10 labeled examples to build rule-based checks, the method achieves training performance comparable to ground truth or rubrics (retaining 88-110% of gains). This significantly lowers the barrier for unsupervised agent self-improvement.

HEAT TRENDHourly heat curve

4 fully observed hours
Now
13.3
Peak
14.415:00
24h change
–

The line compares only sources observed throughout; gap hours are interpolated to keep the trend continuous.

TIMELINECoverage timeline

Total 1 reports · Latest first
  1. Rohan Paul (@rohanpaul_ai)T2·78 pts
    • The Double Ratchet framework allows graders to co-evolve with agents without pre-defined complex rubrics.
    • Effective rule-based graders can be built from just 10 labeled samples, performing well in code and SQL tasks.
    • Experimental data shows the method retains 88% to 110% of the performance gains compared to using ground truth or human rubrics.
AWS and HSBC Introduce Double Ratchet: Self-Evolving Graders Enable Unsupervised Agent Improvement from 10 Samples | Today for AI