A tape-centered LLM agent framework whose structured session logs double as resumable state and training text, demonstrated by distilling a 405B-parameter teacher into an 8B student that matches its score at roughly 1/300th of the cost.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AI 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
TapeAgents: a Holistic Framework for Agent Development and Optimization
A tape-centered LLM agent framework whose structured session logs double as resumable state and training text, demonstrated by distilling a 405B-parameter teacher into an 8B student that matches its score at roughly 1/300th of the cost.