Auto Efficiency Learning Pushing The

---
name: auto-efficiency-learning-pushing-the
description: "The paper 'When EOS Tokens Disagree: Understanding Length Inflation in On-Policy Distillation' revea"
auto_generated: true
created: 2026-09-19
source_insight: "Efficiency learning (Pushing the Limits of LLM Quantization via the Linearity Theorem): How do agent"
status: draft
fitness_score: 0
trials: 0
wins: 0
---

# Auto Efficiency Learning Pushing The

## Trigger
Use when the agent encounters: The paper "When EOS Tokens Disagree: Understanding Length Inflation in On-Policy Distillation" revea

## Verification
- [ ] The skill produces the expected output for its domain
- [ ] No errors in execution
- [ ] Insight quality: actionable and specific

## Notes
Auto-generated from internet insight (Knowledge-to-Action pipeline).
Darwin will trial this skill; it gets promoted only if it wins arena fights.