Preprint

CoACT: Action-Preserving Observation Compression for Coding Agents

Hao Chen Yuancheng Zhu Yitong Zhang Jia Li

Jul 2026 · 1 citation · 43 references

Computer Science

TL;DR

CoACT is built on next-action preservation (NAP), which requires a compressed observation to induce the same next action as the raw observation, and uses an action-preservation reward based on NAP to filter out candidates that would change the agent's next action.

Abstract

LLM-based coding agents solve software-engineering tasks through iterative interactions with development environments, where returned observations accumulate in the context and become a major source of inference cost. Observation compression reduces this cost by shortening observations before they are appended to the context. However, existing methods still exhibit an unsatisfactory efficiency-effectiveness trade-off, as they do not explicitly model how compression affects the agent's subsequent behavior. This paper proposes CoACT, an action-preserving observation compression method for coding agents. CoACT is built on next-action preservation (NAP), which requires a compressed observation to induce the same next action as the raw observation. By checking the agent's immediate next action, NAP provides a practical signal for whether a compression preserves the information needed for continued task solving. During training, a teacher model first generates multiple compressed candidates of each observation. CoACT then uses an action-preservation reward based on NAP to filter out candidates that would change the agent's next action, and uses a length-reduction reward to choose compact candidates as supervision for a lightweight compressor. Experiments on SWE-bench Verified with three agentic models show that CoACT reduces average total token consumption by 33.0% while maintaining task-solving effectiveness close to the uncompressed agent.

arXiv Semantic Scholar

View source

Similar papers

Preprint Aug 2026

EviSD: Evidence-Conditioned Self-Distillation for Search-Augmented Agents

EviSD is proposed, an evidence-conditioned self-distillation framework that uses instance-level supporting evidence as privileged information for search actions and golden answers as complementary privilege for answer actions, and achieves the highest macro-average Exact Match in all evaluated settings.

Jianan Xie, Xin Sun, Zhongqi Chen et al. · 1 citation

Preprint Jul 2026

Experience Memory Graph: One-Shot Error Correction for Agents

Experience Memory Graph is proposed, a framework that reformulates agent failure recovery as a graph matching problem and consistently outperforms state-of-the-art reflection baselines in success rate and average reward, while requiring no test-time trial-and-error.

Wenjun Wang, Yuchen Fang, Fengrui Liu et al. · 0 citations

#machine learning Preprint Aug 2026

Learning Generalizable Behaviors for Terminal Agents

River, a simple training recipe that improves reward quality by filtering low-quality environments and augmenting outcome rewards with process-level behavior regularization is proposed, which achieves the best performance among evaluated open-source RL-trained 8B models across four terminal-agent benchmarks.

Yi-Fan Yao, Bo Pang, Xuan-Phi Nguyen et al. · 2 citations

Preprint Jul 2026

SCATE: Learning to Supervise Coding Agents for Cost-Effective Test Generation

This work proposes SCATE, a framework for adaptive, automated supervision of coding agents that replaces human intervention during test generation by formulating supervision as a contextual bandit problem, which consistently outperforms state-of-the-art non-agentic approaches across all metrics.

Sijia Gu, Noor Nashid, Ali Mesbah · 0 citations

Jul 2026

Function-Aware Fill-in-the-Middle as Mid-Training for Coding Agent Foundation Models

Beyond in-domain gains, mid-training mitigates the capability erosion that agentic post-training otherwise inflicts on non-agent coding and non-coding tool-use benchmarks (tau-bench, BFCL): although the mid-training corpus contains Python code only, the function-call inductive bias survives post-training and yields con...

Yubo Wang, Jiarong Liang, Yuxuan Zhang et al. · arXiv.org · 1 citation

#robotics Preprint Aug 2026

WAM-OPD: On-Policy Distillation for World Action Models

WAM-OPD, a deployment-consistent post-training recipe for a video-first WAM, is introduced and it is suggested that dense teacher supervision on student-induced histories is a promising post-training interface for video-first WAMs.

Liu Yang, Zhuang Jiang, Chenchao Sheng et al. · 2 citations