preprint · arXiv (Cornell University) · 2026
Trained Agentic Context Management
Cited by 0Open access
Abstract
We study long context language models. Instead of training long context natively, or designing a long context harness, we train a model over the simplest possible harness: a tool to call itself with any specified prompt and a tool to read tokens in a range from the input context. We finetune Qwen3.6-35B-A3B on a diverse synthetic dataset using this harness. With only 8,000 tokens of context, our small model is as strong as GPT-5.4 with 1M tokens of context on the OOLONG-synth benchmark when document length exceeds 40K tokens.
Cite this paper
Sandlund, B. (2026). Trained Agentic Context Management. arXiv (Cornell University). https://doi.org/10.48550/arxiv.2610.02404
Sandlund, Bryce. “Trained Agentic Context Management.” arXiv (Cornell University), 2026, https://doi.org/10.48550/arxiv.2610.02404.
Sandlund, Bryce. 2026. “Trained Agentic Context Management.” arXiv (Cornell University). https://doi.org/10.48550/arxiv.2610.02404.
Sandlund, B. (2026) ‘Trained Agentic Context Management’, arXiv (Cornell University). Available at: https://doi.org/10.48550/arxiv.2610.02404.
@misc{sandlund2026trained,
title = {{Trained Agentic Context Management}},
author = {Sandlund, Bryce},
journal = {arXiv (Cornell University)},
year = {2026},
doi = {10.48550/arxiv.2610.02404}
}Made from the index record. Check names, capitalisation and page numbers against the paper. Need another style? AI for Research has over 10,000.
Related papers
How to Train Long-Context Language Models (Effectively)
Tianyu Gao, Alexander Wettig, Howard Yen et al. · Proceedings of the conference - Association for Computational Linguistics. Meeting · 2025
Cited by 132Open access
Bootstrap Your Own Context Length
Liang Wang, Nan Yang, Xingxing Zhang et al. · arXiv (Cornell University) · 2024
Cited by 10Open access
Scaling Context, Not Parameters: Training a Compact 7B Language Model for Efficient Long-Context Processing
Chen Henry Wu, Yin Song · Proceedings of the conference - Association for Computational Linguistics. Meeting · 2025
Cited by 2Open access
From 128K to 4M: Efficient Training of Ultra-Long Context Large Language Models
Chejian Xu, Wei Ping, Peng Xu et al. · Findings of the Association for Computational Linguistics: ACL (Findings of ACL) · 2026
Cited by 18Open access
Scaling Instruction-Tuned LLMs to Million-Token Contexts via Hierarchical Synthetic Data Generation
He, Linda, Jue Wang, Maurice Weber et al. · arXiv (Cornell University) · 2025
Cited by 4Open access
ACC: Compiling Agent Trajectories for Long-Context Training
Qisheng Su, Zhen Fang, Shiting Huang et al. · arXiv (Cornell University) · 2026
Cited by 5Open access
OctoLong: Mid-Training On Cross-Repository Code Contexts Enhances Long-Context Modeling
Indraneil Paul, Falko Helm, Goran Glavaš et al. · TUbilio (Technical University of Darmstadt) · 2026
Cited by 2Open access
OPSDL: On-Policy Self-Distillation for Long-Context Language Models
Xinsen Zhang, Zhenkai Ding, Tianjun Pan et al. · arXiv (Cornell University) · 2026
Cited by 33Open access