← All concepts

self improvement without retraining

1 articles · 7 co-occurring · 0 contradictions · 0 briefs

Demonstrates LLMs can improve through context adaptation (generation-reflection-curation) without weight updates or labeled data

Related concepts
rag retrieval 1 prompt optimization 1 prompt engineering 1 multi agent orchestration 1 context window management 1 context collapse 1 agent memory 1
Evidence chain (1 articles, showing 1)
Agentic Context Engineering: Evolving Contexts for Self-Improving Language Models | UC Berkeley School of Information supports

Demonstrates LLMs can improve through context adaptation (generation-reflection-curation) without weight updates or labeled data

query this concept
$ db.articles("self-improvement-without-retraining")
$ db.cooccurrence("self-improvement-without-retraining")
$ db.contradictions("self-improvement-without-retraining")