Coding agents re-send large file reads and tool outputs to a frontier LLM every turn, and this context dominates their token bill. General-purpose prompt compressors are trained on prose and suit code...
Paritok-4B is a 4‑billion‑parameter LoRA compressor designed for coding agents, which extracts and retains key spans of code rather than paraphrasing them. It is intent‑conditioned, selecting lines that are most relevant to the agent’s current task, and achieves high fidelity with 96% of identifiers, paths, and numbers preserved. Trained on 67,074 real OpenHands trajectories and fine‑tuned on Qwen3‑4B, it compresses agent context to about 25.7% of its original size while keeping 86.5% of the uncompressed solve quality across 300 SWE‑bench Lite instances.
By Jiayu Shi, Luzhuo Chen
arXiv:2610.15093v1 Announce Type: cross
Abstract: AI coding agents operate in iterative loops: they modify code, run tools, inspect the results, and decide what to do next. We call the design of this...
By Pramod Kumar, Raj Kumar Pandey
The paper introduces LOHA, a context layout that compresses older tool observations into soft tokens while keeping the agent’s own turns and the last K observations in plain text, and ACD, a training method that distills full‑text predictions into this latent representation while anchoring behavior on plain text. This approach reduces context per call by up to 57% without significant loss in resolve rates, and improves instance throughput in single‑GPU serving. Experiments on SWE‑bench Verified show that K=3 yields a 43–57% compression with only modest performance impact, while larger windows favor task performance over compression.
By Zhensheng Zou (Peking University), Guoqing Wang (Peking University), Dan Hao (Peking University)
KuaFu is a unified behavior‑compression layer that reduces each user behavior item to 2–4 tokens, dramatically shrinking per‑item cache size while preserving fidelity through a four‑stage training process. In production across four profiling tasks, it matches or outperforms uncompressed single‑task models, boosts GPU throughput by 37–350%, and saves 190 GPUs. On public benchmarks it consistently beats prior compressors at the same compression ratio, and on RecBench a 4B KuaFu model outperforms its 8B counterpart by 1.90 points, contributing to a 1.37% lift in overall GMV on Tencent’s advertising and recommendation platform.
By Jiahao Hui, Lin Zhu, Yishen Hu, Jingdong Shu, Zetai Jiang, Xining Ran, Ben Tan, Yeshou Cai, Gong Chen, Haijie Gu, Jie Jiang
arXiv:2610.34763v1 Announce Type: cross
Abstract: Coding agents resend a growing context on every model call, and context policies are usually judged by the prompt tokens they save. Executed work als...
By Zhongming Yu, Yitong Ma, Hengjia Yu, Boqin Yuan, Xiangye Lin, Hejia Zhang, Jishen Zhao
arXiv:2607. 15516v1 Announce Type: cross Abstract: Production LLM deployments combine two cost-reduction primitives: prompt caching (a discounted rate for re-used token prefixes) and prompt compression (fewer tokens sent).
By Yan Song
arXiv:2610.08811v1 Announce Type: new
Abstract: As context windows scale to tens or hundreds of thousands of tokens, KV cache compression has become essential for efficient LLM inference. Existing me...
By Linfeng Dong
arXiv:2609.23790v1 Announce Type: new
Abstract: Every node in a multi-agent large language model (LLM) workflow retrieves context from memory and injects it into its prompt, where those injected toke...
By Vivek Kumar Singh, Preeti Priyam, Gautam Bhowmick
arXiv:2608.29363v1 Announce Type: new
Abstract: Enterprise data agents answer business queries by chaining many tool calls over multiple reasoning steps, routinely accumulating hundreds of thousands...
By Ziqi Lin, Ye Wu, Mengying Yang, Xu Liu, Yizhou Liu, Qiang Ke, Qin Guo
arXiv:2608. 16370v1 Announce Type: new Abstract: Task completion is the standard metric for evaluating context compression, yet it is incomplete: compression can increase an agent's interaction cost by forcing it to reacquire dropped state while leaving completion statistically unchanged.
By Shuyu Liu
Diffusion language models (dLLMs) generate text by iteratively denoising a masked response and can commit multiple output positions per model invocation. Their bidirectional attention prevents exact autoregressive-style KV caching, since committing one position shifts the KV activations of all others.