Skip to content
NewA calmer, faster ZeroHour

AREX-2: Advancing Self-Improving Agents through Long-Horizon Reflective Tasks

Hugging Face daily papersby Hongjin Qian, Chaofan Li, Kun Luo ()1 min read3
AI summary · grok-4.7Summaries by Vani AI

AREX-2 trains a Qwen3.8-27B agent on long-horizon reflective tasks, lifting research and coding benchmarks.

AREX-2 synthesizes long-horizon improvement trajectories from machine-learning and algorithmic programming tasks to teach reflection and sustained test-time iteration. An agent built on Qwen3.8-27B scores 81.8 on MLE-bench Lite and 70.7 on Frontier-CS. It transfers to deep research with 84.0 on BrowseComp, 52.6 on HLE, 92.2 on GAIA, and 93.8 on DeepSearchQA, and keeps improving as its round budget grows.

Key facts
  • Training uses verifiable ML and programming improvement trajectories.
  • Qwen3.8-27B scores 81.8 on MLE-bench Lite and 70.7 on Frontier-CS.
  • Transfers to BrowseComp 84.0, HLE 52.6, GAIA 92.2, and DeepSearchQA 93.8.
  • Results keep improving as the iteration budget grows.
Explain this AI news with Vani AIWhat was announced, what is new and why it matters, in a new chat

From the original article

1 min read at huggingface.co

We present AREX-2, an effort to advance the self-improving capability of LLM agents, which we define as the ability to iteratively refine a solution at test time. This ability rests on two complementary capabilities: reflection, which produces a solution better than the current one, and long-horizon execution, which keeps the iteration effective over many rounds.

Hugging Face daily papers

By Hongjin Qian, Chaofan Li, Kun Luo · huggingface.co

Read the full article at Hugging Face daily papers

A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.