Can Language Models Solve Olympiad Programming?
This work addresses the challenge of assessing language models' capabilities in complex algorithmic reasoning for researchers and practitioners, though it is incremental as it builds on existing LM methods.
The paper tackles the problem of evaluating language models on competitive programming tasks by introducing the USACO benchmark with 307 problems, finding that GPT-4 achieves only 8.7% accuracy with zero-shot prompting, and improving it to 20.2% with advanced inference methods, but this remains far from solving the benchmark.
Computing olympiads contain some of the most challenging problems for humans, requiring complex algorithmic reasoning, puzzle solving, in addition to generating efficient code. However, it has been understudied as a domain to evaluate language models (LMs). In this paper, we introduce the USACO benchmark with 307 problems from the USA Computing Olympiad, along with high-quality unit tests, reference code, and official analyses for each problem. These resources enable us to construct and test a range of LM inference methods for competitive programming for the first time. We find GPT-4 only achieves a 8.7% pass@1 accuracy with zero-shot chain-of-thought prompting, and our best inference method improves it to 20.2% using a combination of self-reflection and retrieval over episodic knowledge. However, this is far from solving the benchmark. To better understand the remaining challenges, we design a novel human-in-the-loop study and surprisingly find that a small number of targeted hints enable GPT-4 to solve 13 out of 15 problems previously unsolvable by any model and method. Our benchmark, baseline methods, quantitative results, and qualitative analysis serve as an initial step toward LMs with grounded, creative, and algorithmic reasoning.