Newest papers

LongZero: Enhancing Long-Context Reasoning in LLMs from Zero Data

Zhennan Wan, Jianfei Chen · 2026 · arXiv (Cornell University)

DOI: 10.48550/arxiv.2610.30132 · cited by 0 works (OpenAlex) · open-access PDF

Indonesian

Pembelajaran penguatan data (RL) adalah cara yang efektif untuk memperkuat pertimbangan konteks panjang dari model bahasa besar (LLM), tetapi membutuhkan data pelatihan yang panjang, beragam, menantang, dan dapat diverifikasi. Anotasi manusia dari data tersebut mahal, data yang terjadi secara alami terlalu sedikit untuk mempertahankan RL, dan metode sintesis yang ada bergantung pada corpora eksternal atau model guru yang lebih kuat, yang menjadi lebih sulit untuk diperoleh karena model eksternal yang dilatih menjadi lebih kuat. Baru-baru ini, RL dari data nol, di mana sebuah model belajar dari tugas-tugas yang diusulkan sendiri, telah mencapai kinerja yang sangat baik, tetapi metode yang ada menargetkan tugas-tugas promotori pendek. Oleh karena itu, kami mengusulkan $textbf{LongZero}$, sebuah metode yang memungkinkan LLM meningkatkan pertim

Machine translation (Meta M2M100); the English original follows.

Abstract

Reinforcement learning (RL) is an effective way to strengthen the long-context reasoning of large language models (LLMs), but it needs training data that are long, diverse, challenging, and verifiable. Human annotation of such data is expensive, naturally occurring data are too scarce to sustain RL, and existing synthesis methods depend on external corpora or stronger teacher models, which become harder to obtain as the trained models get stronger. Recently, RL from zero data, in which a model learns from tasks it proposes itself, has achieved excellent performance, but existing methods target short-context tasks. Therefore, we propose $\textbf{LongZero}$, a method that lets an LLM improve its own long-context reasoning without any external data. Instead of writing questions and answers over existing documents, LongZero first has the model synthesize the question and answer with its supporting evidence in a compact structured form, and only then separately expands the evidence into a long multi-document context. LongZero combines these self-written multi-hop reasoning tasks with self-written multi-round retrieval tasks and trains the model with GRPO. We evaluate LongZero on two LLMs across five long-context reasoning benchmarks and compare it with RL on data synthesized from external corpora. Experimental results show that LongZero can generalize to real tasks, and outperforms both baselines, suggesting a viable and promising path to improving long-context reasoning of LLMs.

Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese

Topics: Computer science, Artificial intelligence, Annotation, Path (computing), Reinforcement learning, Zero (linguistics)

Sources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange