跳到正文
HuggingFace Daily Papers(社區熱門論文)·· 2026-10-04AI 評分38

CURIO: Curiosity-Driven Test-Time Learning for Open-Ended Discovery

CURIO: Curiosity-Driven Test-Time Learning for Open-Ended Discovery

AI 導讀

CURIO 提出好奇心驅動的測試時學習框架,透過內在好奇心世界模型補充任務回饋。在 Qwen3 8B 至 235B 模型上,六項數學探索任務中有五項優於僅用任務回饋的強化學習對照,Circle Packing 達到最佳已報告表現;單細胞去噪在各測試規模的兩個留出語料上,去噪 Score 和均方誤差均有改善。相對提升在 Hadamard 達 18.3%,去噪 Score 達 10.8%。

正文

Published on Oct 4

·

Submitted by

Fangxu

on Oct 14

Authors:

,

,

,

,

,

,

Abstract

Open-ended discovery requires learning from repeated attempts while continuing to explore directions whose value is not yet apparent. Search with a frozen large language model (LLM) can reuse previous solutions in context, but cannot update the model from its successes and failures on the test problem. Reinforcement learning (RL) enables such adaptation; however, strongly favoring high-reward trajectories may suppress low-reward yet potentially promising directions too early. We introduce CURIO, a curiosity-driven test-time learning framework that complements task feedback with an Intrinsic Curiosity World Model (ICWM). The ICWM learns transitions in the policy's hidden-state representation and supplies prediction-error bonuses at sampled tokens outside the policy's top-k choices. Epoch normalization and an annealed weight regulate their contribution to the policy update. On six mathematical discovery tasks and single-cell denoising with Qwen3 backbones from 8B to 235B, three-run means improve over a matched task-only RL control on five mathematical objectives, match the best reported performance on Circle Packing, and improve denoising Score and mean squared error (MSE) on both held-out corpora at every tested scale. Relative gains reach 18.3% on Hadamard and 10.8% on denoising Score. Code-diversity measurements show greater structural variation among generated programs, supporting curiosity as a complementary exploration signal for learning in open-ended discovery.

View arXiv page View PDF Project page GitHub Add to collection

Get this paper in your agent:

hf papers read 2610.04851

Don't have the latest CLI?

curl -LsSf https://hf.co/cli/install.sh | bash

Models citing this paper 0

No model linking this paper

Cite arxiv.org/abs/2610.04851 in a model README.md to link it from this page.

Datasets citing this paper 0

No dataset linking this paper

Cite arxiv.org/abs/2610.04851 in a dataset README.md to link it from this page.

Spaces citing this paper 0

No Space linking this paper

Cite arxiv.org/abs/2610.04851 in a Space README.md to link it from this page.

Collections including this paper 0

No Collection including this paper

Add this paper to a collection to link it from this page.

來源:HuggingFace Daily Papers(社區熱門論文) · huggingface.co