Timezone: »

 
Language Models Can Teach Themselves to Program Better
Patrick Haluptzok · Matthew Bowers · Adam Kalai

@
in Deep Reinforcement Learning Workshop »
Event URL: https://openreview.net/forum?id=_5BZwkZRFc9 »

Recent Language Models (LMs) achieve breakthrough performance in code generation when trained on human-authored problems, even solving some competitive-programming problems. Self-play has proven useful in games such as Go, and thus it is natural to ask whether LMs can generate their own instructive programming problems to improve their performance. We show that it is possible for an LM to synthesize programming problems and solutions, which are filtered for correctness by a Python interpreter. The LM’s performance is then seen to improve when it is fine-tuned on its own synthetic problems and verified solutions; thus the model “improves itself” using the Python interpreter. Problems are specified formally as programming puzzles [Schuster et al., 2021], a code-based problem format where solutions can easily be verified for correctness by execution. In experiments on publicly-available LMs, test accuracy more than doubles. This RL approach demonstrates the potential for code LMs, with an interpreter, to generate instructive problems and improve their own performance.

[ Poster]  [ OpenReview]  [ Topia] 

Author Information

Patrick Haluptzok
Matthew Bowers (Massachusetts Institute of Technology)
Matthew Bowers

PhD student at MIT co-advised by Armando Solar-Lezama in EECS and Josh Tenenbaum in BCS. Research in combining methods from programming languages (PL) research with machine learning to tackle problems in artificial intelligence.

Adam Kalai (Microsoft Research New England (-(-_(-_-)_-)-))

More from the Same Authors