95 papers
DIRECTOR: Dynamic Index-based Recommendation with Transport-Optimized Retrieval
Yuanhao Pu, Chenghao Zhang, Chao Feng +2
The paper introduces DIRECTOR, a parallel reranking framework that uses dynamic retrieval indices and entropy‑regularized optimal transport to generate duplicate‑free recommendatio…
Rethinking Heterogeneous LLM Merging: A Weighted Model Averaging Perspective
Jiahe Fan, Yinghao Hou, Si Chen +3
Can large language models with substantially different parameter spaces be merged by direct weighted averaging, without training or semantic alignment? Existing heterogeneous fusio…
CR: Cross-sample Consistency Regularization Mitigates Feature Splitting and Absorption in Sparse Autoencoders
Haoran Jin, Xiting Wang, Shijie Ren +2
Sparse Autoencoders (SAEs) are widely used to interpret large language models by decomposing activations into sparse, human-understandable features, but scaling to large dictionari…
RAVEN: A Regime-Aware Variable-context Expert Network for Financial Time Series Forecasting
Cheng He, Zhenyu Guan, Xijie Liang +6
Financial time series forecasting presents structural challenges absent from standard benchmarks. Log-returns are non-stationary, exhibit exceptionally low signal-to-noise (SNR) ra…
EquivPruner: Boosting Efficiency and Quality in LLM-Based Search via Action Pruning
Jiawei Liu, Qisi Chen, Jianshu Zhang +2
Large Language Models (LLMs) excel at complex reasoning through search algorithms, yet current strategies often suffer from massive token consumption due to redundant exploration o…
Large-Scale OD Matrix Estimation with A Deep Learning Method
Zheli Xiong, Defu Lian, Enhong Chen +2
The estimation of origin-destination (OD) matrices is a crucial aspect of Intelligent Transport Systems (ITS). It involves adjusting an initial OD matrix by regressing the current…