paper-with-me

홈 › Papers

Roll the dice & look before you leap: Going beyond the creative limits of next-token prediction

2025-04-21 · Vaishnavh Nagarajan, Chen Henry Wu, Charles Ding, aditi raghunathan

We design a suite of minimal algorithmic tasks that are a loose abstraction of open-ended real-world tasks. This allows us to cleanly and controllably quantify the creative limits of the present-day language model. Much like real-world tasks that require a creative, far-sighted leap of thought, our tasks require an implicit, open-ended stochastic planning step that either (a) discovers new connections in an abstract knowledge graph (like in wordplay, drawing analogies, or research) or (b) constructs new patterns (like in designing math problems or new proteins). In these tasks, we empirically and conceptually argue how next-token learning is myopic and memorizes excessively; comparatively, multi-token approaches, namely teacherless training and diffusion models, excel in producing diverse and original output. Secondly, in our tasks, we find that to elicit randomness from the Transformer without hurting coherence, it is better to inject noise right at the input layer (via a method we dub hash-conditioning) rather than defer to temperature sampling from the output layer. Thus, our work offers a principled, minimal test-bed for analyzing open-ended creative skills, and offers new arguments for going beyond next-token learning and softmax-based sampling. We make part of the code available under https://github.com/chenwu98/algorithmic-creativity

📄 PDF Abstract BibTeX arXiv:2504.15266

Code (1)

chenwu98/algorithmic-creativity 공식 구현 pytorch

Tasks

Math

Methods 이 논문이 사용한 방법론

Linear Layer A Linear Layer is a projection $\mathbf{XW + b}$.
Multi-Head Attention 설명 없음
Dense Connections Dense Connections, or Fully Connected Connections, are a type of layer in a deep neural network that use a linear operation where every input is connected to every output…
Adam 설명 없음
Attention 설명 없음
Dropout Dropout is a regularization technique for neural networks that drops a unit (along with connections) at training time with a specified probability $p$ (a common value is…
Diffusion Diffusion models generate samples by gradually removing noise from a signal, and their training objective can be expressed as a reweighted variational lower-bound…
Layer Normalization Unlike batch normalization, Layer Normalization directly estimates the normalization statistics from the summed inputs…

Similar Papers 제목 키워드 기반

Look Before You Leap: Learning Landmark Features for One-Stage Visual Grounding

2021-04-09 · CVPR 2021 1 · Binbin Huang, Dongze Lian, Weixin Luo, Shenghua Gao

An LBYL (`Look Before You Leap') Network is proposed for end-to-end trainable one-stage visual grounding. The idea behind LBYL-Net is intuitive and straightforward: we follow a language's description to localize the targ…

DescriptiveObjectVisual Grounding

Look Before You Leap: Improving Text-based Person Retrieval by Learning A Consistent Cross-modal Common Manifold

2022-09-13 · Zijie Wang, Aichun Zhu, Jingyi Xue, Xili Wan 외

The core problem of text-based person retrieval is how to bridge the heterogeneous gap between multi-modal data. Many previous approaches contrive to learning a latent common manifold mapping paradigm following a \textbf…

Person RetrievalRetrievalText based Person RetrievalText-based Person Retrieval

Can a Small Model Learn to Look Before It Leaps? Dynamic Learning and Proactive Correction for Hallucination Detection

2025-11-08 · Zepeng Bao, Shen Zhou, Qiankun Pi, Jianhao Chen 외 arxiv

Hallucination in large language models (LLMs) remains a critical barrier to their safe deployment. For hallucination detection to be practical in real-world scenarios, the use of efficient small models is essential to en…

Look Before You Leap: Autonomous Exploration for LLM Agents

2026-05-15 · Ziang Ye, Wentao Shi, Yuxin Liu, Yu Wang 외 arxiv

Large language model based agents often fail in unfamiliar environments due to premature exploitation: a tendency to act on prior knowledge before acquiring sufficient environment-specific information. We identify autono…

Reinforcement Learning

Look Before You Leap: Safe Model-Based Reinforcement Learning with Human Intervention

2021-11-10 · Yunkun Xu, Zhenyu Liu, Guifang Duan, Jiangcheng Zhu 외

Safety has become one of the main challenges of applying deep reinforcement learning to real world systems. Currently, the incorporation of external knowledge such as human oversight is the only means to prevent the agen…

BlockingDecision MakingDeep Reinforcement LearningModel-based Reinforcement Learning+3