Reasoning in Token Economies: Budget-Aware Evaluation of LLM Reasoning Strategies
This addresses the efficiency and fairness in benchmarking LLM methods for researchers and practitioners, highlighting an incremental but important oversight in current evaluation practices.
The paper tackles the problem of evaluating LLM reasoning strategies by showing that traditional performance-focused evaluations are skewed because they ignore compute costs, and it finds that simpler baselines like chain-of-thought self-consistency often outperform complex strategies when given comparable compute resources, with some strategies even degrading with more budget.
A diverse array of reasoning strategies has been proposed to elicit the capabilities of large language models. However, in this paper, we point out that traditional evaluations which focus solely on performance metrics miss a key factor: the increased effectiveness due to additional compute. By overlooking this aspect, a skewed view of strategy efficiency is often presented. This paper introduces a framework that incorporates the compute budget into the evaluation, providing a more informative comparison that takes into account both performance metrics and computational cost. In this budget-aware perspective, we find that complex reasoning strategies often don't surpass simpler baselines purely due to algorithmic ingenuity, but rather due to the larger computational resources allocated. When we provide a simple baseline like chain-of-thought self-consistency with comparable compute resources, it frequently outperforms reasoning strategies proposed in the literature. In this scale-aware perspective, we find that unlike self-consistency, certain strategies such as multi-agent debate or Reflexion can become worse if more compute budget is utilized.