paper-with-me

Papers

Decoding in Geometry: Alleviating Embedding-Space Crowding for Complex Reasoning

2026-01-30 · Yixin Yang, Qingxiu Dong, Zhifang Sui arxiv

Sampling-based decoding underlies complex reasoning in large language models (LLMs), where decoding strategies critically shape model behavior. Temperature- and truncation-based methods reshape the next-token distribution through global probability reweighting or thresholding to balance the quality-diversity tradeoff. However, they operate solely on token probabilities, ignoring fine-grained relationships among tokens in the embedding space. We uncover a novel phenomenon, embedding-space crowding, where the next-token distribution concentrates its probability mass on geometrically close tokens in the embedding space. We quantify crowding at multiple granularities and find a statistical association with reasoning success in mathematical problem solving. Motivated by this finding, we propose CraEG, a plug-and-play sampling method that mitigates crowding through geometry-guided reweighting. CraEG is training-free, single-pass, and compatible with standard sampling strategies. Experiments on multiple models and benchmarks demonstrate improved generation performance, with gains in robustness and diversity metrics.

📄 PDF Abstract BibTeX arXiv:2601.22536

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

Doubly Stochastic Neighbor Embedding on Spheres

2016-09-07 · Yao Lu, Jukka Corander, Zhirong Yang

Stochastic Neighbor Embedding (SNE) methods minimize the divergence between the similarity matrix of a high-dimensional data set and its counterpart from a low-dimensional embedding, leading to widely applied tools for d…

Data Visualization

Over-Squashing in Riemannian Graph Neural Networks

2023-11-27 · Julia Balla

Most graph neural networks (GNNs) are prone to the phenomenon of over-squashing in which node features become insensitive to information from distant nodes in the graph. Recent works have shown that the topology of the g…

SpaceMAP: Visualizing Any Data in 2-dimension by Space Expansion

2021-09-29 · Xinrui Zu, Qian Tao

Dimensionality reduction (DR) and visualization of high-dimensional data is of theoretical and practical value in machine learning and related fields. In theory, there exists an intriguing, non-intuitive discrepancy betw…

Dimensionality Reduction

Home-made blues: Residential crowding and mental health in Beijing, China

2022-07-16 · Xize Wang, Tao Liu

Although residential crowding has many well-being implications, its connection to mental health is yet to be widely examined. Using survey data from 1613 residents in Beijing, China, we find that living in a crowded plac…

Geometry-Aware Decoding with Wasserstein-Regularized Truncation and Mass Penalties for Large Language Models

2026-02-10 · Arash Gholami Davoodi, Navid Rezazadeh, Seyed Pouyan Mousavi Davoudi, Pouya Pezeshkpour arxiv

Large language models (LLMs) must balance diversity and creativity against logical coherence in open-ended generation. Existing truncation-based samplers are effective but largely heuristic, relying mainly on probability…