learn

To correct inaccurate judgments and calibrate scores across sessions, you provide targeted feedback by marking outputs as correct, incorrect, or partial. This applies corrections, explains errors, and adjusts accuracy using brain_cynic_feedback for updates and brain_learning for state checks.

The best way to provide feedback on a judgment is to specify whether it was correct, incorrect, or partial, and attach an explanation. You can also suggest corrected scores and rationale to trigger learning weight calibration and improve future performance.

Yes, you can review learned patterns and systematic errors from past judgments. By performing state checks via brain_learning, you can identify systematic errors and calibrate biases to improve overall accuracy across cross-session learning.

Cross-session learning works by collecting your feedback on judgments and applying it to trigger and calibrate learning weights. It uses brain_cynic_feedback for updates and brain_learning for state checks, ensuring systematic errors are fixed and scores are calibrated over time.

Providing a rationale when marking a judgment as incorrect or partial is highly recommended. Attaching explanations and suggesting corrected scores enables the system to explain errors, calibrate biases, and effectively learn from mistakes across sessions.

You should run state checks on brain_learning to calibrate judgment accuracy whenever you have applied new feedback via brain_cynic_feedback. This allows you to review the learning state, identify systematic errors, and ensure biases are properly calibrated.