# Token-Hungry, Yet Precise: DeepSeek R1 Highlights the Need for Multi-Step Reasoning Over Speed in MATH

arXiv: https://arxiv.org/abs/2501.18576 · HTML: https://mlsanity.com/paper/2501.18576/

## Key facts

- Cited by 10 papers in the map (data through 14 Oct 2026).
- Placed in Reasoning benchmarks and evaluation, one of 17 directions in LLM reasoning.
- Builds on 4 of its 11 references that are in the map.
- First posted on arXiv 30 Jan 2025.

## Placed in

- LLM reasoning / [Reasoning benchmarks and evaluation](https://mlsanity.com/lr/reasoning-benchmarks-and-evaluation/)

## Builds on (most cited first, 4 of 4)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [Training Verifiers to Solve Math Word Problems](https://mlsanity.com/paper/2110.14168/) (2021)
- [Measuring Mathematical Problem Solving With the MATH Dataset](https://mlsanity.com/paper/2103.03874/) (2021)
- [Token-by-Token Regeneration and Domain Biases: A Benchmark of LLMs on Advanced Mathematical Problem-Solving](https://mlsanity.com/paper/2501.17084/) (2025)

## Built on by (most cited first, 10 of 10)

- [A Comparison of DeepSeek and Other LLMs](https://mlsanity.com/paper/2502.03688/) (2025)
- [Evaluating Mathematical Reasoning Across Large Language Models: A Fine-Grained Approach](https://mlsanity.com/paper/2503.10573/) (2025)
- [Benchmarking LLM for Code Smells Detection: OpenAI GPT-4.0 vs DeepSeek-V3](https://mlsanity.com/paper/2504.16027/) (2025)
- [GUI-Rise: Structured Reasoning and History Summarization for GUI Navigation](https://mlsanity.com/paper/2510.27210/) (2025)
- [SE-GA: Memory-Augmented Self-Evolution for GUI Agents](https://mlsanity.com/paper/2605.16883/) (2026)
- [Reasoner for Real-World Event Detection: Scaling Reinforcement Learning via Adaptive Perplexity-Aware Sampling Strategy](https://mlsanity.com/paper/2507.01327/) (2025)
- [Counting Cycles with AI: Counting Cycles with AI: Computationally Efficient Equivalent Forms with Applications](https://mlsanity.com/paper/2505.17964/) (2025)
- [A Showdown of ChatGPT vs DeepSeek in Solving Programming Tasks](https://mlsanity.com/paper/2503.13549/) (2025)
- [Optimizing Humor Generation in Large Language Models: Temperature Configurations and Architectural Trade-offs](https://mlsanity.com/paper/2504.02858/) (2025)
- [CACR:Reinforcing Temporal Answer Grounding in Instructional Video via Candidate-Aware Causal Reasoning](https://mlsanity.com/paper/2606.08436/) (2026)

## Abstract (arXiv)

This study investigates the performance of the DeepSeek R1 language model on 30 challenging mathematical problems derived from the MATH dataset, problems that previously proved unsolvable by other models under time constraints. Unlike prior work, this research removes time limitations to explore whether DeepSeek R1's architecture, known for its reliance on token-based reasoning, can achieve accurate solutions through a multi-step process. The study compares DeepSeek R1 with four other models (gemini-1.5-flash-8b, gpt-4o-mini-2024-07-18, llama3.1:8b, and mistral-8b-latest) across 11 temperature settings. Results demonstrate that DeepSeek R1 achieves superior accuracy on these complex problems but generates significantly more tokens than other models, confirming its token-intensive approach. The findings highlight a trade-off between accuracy and efficiency in mathematical problem-solving with large language models: while DeepSeek R1 excels in accuracy, its reliance on extensive token generation may not be optimal for applications requiring rapid responses. The study underscores the importance of considering task-specific requirements when selecting an LLM and emphasizes the role of temperature settings in optimizing performance.