Papers
arxiv:2509.25176

SIRI: Scaling Iterative Reinforcement Learning with Interleaved Compression

Published on Sep 29
· Submitted by Yushi Bai on Sep 30
Authors:
,
,
,

Abstract

SIRI, a reinforcement learning approach with interleaved compression and expansion, enhances the efficiency and accuracy of large reasoning models by dynamically adjusting the reasoning budget.

AI-generated summary

We introduce SIRI, Scaling Iterative Reinforcement Learning with Interleaved Compression, a simple yet effective RL approach for Large Reasoning Models (LRMs) that enables more efficient and accurate reasoning. Existing studies have observed repetitive thinking patterns in LRMs, and attempts to reduce them often come at the cost of performance. In this paper, we show that this trade-off can be overcome through a training regime that iteratively alternates between compressing and expanding the reasoning budget, by dynamically adjusting the maximum rollout length during training. The compression phase cuts the rollout length, forcing the model to make precise and valuable decisions within a limited context, which effectively reduces redundant tokens and increases reasoning density. The expansion phase then relaxes the length limit, providing space for the model to explore and plan in long-horizon settings. Remarkably, we find that after each compression-expansion cycle, the model's performance improves even as its output length decreases, steadily pushing it closer to the Pareto frontier in the performance-efficiency trade-off. Training on DeepSeek-R1-Distill-Qwen-1.5B, SIRI-low improves performance on AIME24 by 43.2% while reducing token usage by 46.9% after three iterations, and SIRI-high achieves the highest accuracy compared to all other methods (Figure 1). Our findings shed light on the potential of periodically oscillating the LRM's output truncation length during training to dynamically balance exploration and efficiency in reasoning, converging towards an optimal "sweet spot" between the two. Our models are publicly available.

Community

Paper submitter

In this paper, we propose SIRI, a simple but effective approach to enhance the performance of
LRMs while pruning repetitive reasoning traces. We apply expansion and compression of the token
budget iteratively, encouraging exploration and consolidation in turn. SIRI models: https://huggingface.co/collections/THU-KEG/siri-68d65a4ecf9f20dac7322dfe

This is an automated message from the Librarian Bot. I found the following papers similar to this paper.

The following papers were recommended by the Semantic Scholar API

Please give a thumbs up to this comment if you found it helpful!

If you want recommendations for any Paper on Hugging Face checkout this Space

You can directly ask Librarian Bot for paper recommendations by tagging it in a comment: @librarian-bot recommend

Sign up or log in to comment

Models citing this paper 4

Datasets citing this paper 0

No dataset linking this paper

Cite arxiv.org/abs/2509.25176 in a dataset README.md to link it from this page.

Spaces citing this paper 0

No Space linking this paper

Cite arxiv.org/abs/2509.25176 in a Space README.md to link it from this page.

Collections including this paper 1