← Back to all news

Retrieval-augmented generation

Retrieval pipelines, vector search, chunking and reranking: how models are grounded in a corpus instead of their weights.

6,299 stories · RSS feed

arXiv Machine Learning
Sep 22

LE4Mob: Towards Inductive, Distance-Aware and General-Purpose Location Embedding for Human Mobility Modelling

LE4Mob is a new location embedding framework that learns inductive, distance‑aware representations from geographic context, enabling it to encode unseen locations and preserve spatial relationships. It builds on contrastive language‑location pre‑training and adds a regularisation objective that encourages the embedding space to reflect geographic distance. Experiments on next‑location prediction and commuter flow generation across multiple datasets show that LE4Mob outperforms strong baselines, especially in inductive settings and when downstream models use direct interactions between location embeddings.

By Xinglei Wang, Stephen Law, Zichao Zeng, Junyuan Liu, Guangsheng Dong, Tao Cheng
rag
More like this →
arXiv Machine Learning
Sep 22

From Latent Biomarkers to Clinical Rules: Embedding-Guided Rule Mining and Attribution-Based Translation for Interpretable Tabular Learning

The paper introduces a four-step pipeline that mines decision rules in the latent space of an FT-Transformer and then translates those rules back into measurable clinical features. By treating embedding dimensions that separate patient groups as latent biomarkers, small decision trees are used to extract rules, which are then mapped to raw features using gradient-input saliency and CLS attention attribution. Across six public clinical datasets, the translated rules generally outperformed raw-feature rules, achieving significant AUROC gains, though some high-performing latent rules could not be fully captured by simple raw-feature conditions.

By Majid Lotfian Delouee, Hamed Ayoobi, Sjors G. J. G. In 't Veld, Martijn C. Schut
llmsragdiffusion
More like this →
arXiv Machine Learning
Sep 22

EmbeddGAN: A Novel GAN Framework Using an Embedding Network and Gini Distance Correlation

EmbeddGAN introduces a new GAN framework that replaces the traditional discriminator with an embedding network trained to maximize statistical dependence between embeddings and real/fake labels using Gini distance correlation (gCor). The generator simultaneously minimizes this dependence, encouraging real and generated samples to become indistinguishable in the learned low‑dimensional embedding space. Experiments on MNIST, CIFAR‑10, and CelebA show competitive performance and notably more stable training dynamics compared to established baselines.

By MaTais Caldwell, Yixin Chen, Xin Dang, Charles Walter
ragdiffusionsafety
More like this →
arXiv Machine Learning
Sep 22

Beyond Similarity: Coverage-Aware Prompt Selection for Time Series Forecasting with LLMs

The paper introduces CASP-LLM, a coverage‑aware semantic prompting framework that mitigates bias toward dominant temporal patterns in prompt‑based time series forecasting. Unlike traditional similarity‑based retrieval that selects top‑K candidates by cosine similarity, CASP‑LLM uses usage‑tracking and a saturating‑gate regularizer to diversify prompt selection without adding learnable parameters. Experiments on six long‑term and the M4 short‑term benchmarks show that CASP‑LLM matches or outperforms similarity‑based LLM forecasters in most settings, with failures traced to cross‑batch usage rather than within‑retrieval redundancy.

By Daeun Ji, Minkyoung Kim, Dongkuk Kim, Yohan Lee, Beomsoo Kim, Beakcheol Jang
llmsragbenchmarkssafety
More like this →
arXiv Machine Learning
Sep 22

Whitening Inverts the Hierarchy: What the Norm of a Whitened Embedding Measures

The paper investigates the use of the squared norm of a whitened foundation‑model embedding as a training‑free likelihood surrogate. It shows that the apparent Gaussianity of whitened coordinates stems from the projection central limit theorem, not from a true joint Gaussian distribution, and that the norm is systematically over‑dispersed compared to a Gaussian reference. The authors explain that whitening reverses the encoder’s spectral hierarchy, concentrating norm contributions in near‑degenerate directions dominated by noise, and propose interpreting the squared norm as a Mahalanobis measure of semantic atypicality rather than a log‑likelihood.

By Mohammed Ahnouch, Lotfi Elaachak
rag
More like this →
arXiv Machine Learning
Sep 22

Rethinking Class Imbalance for Single-Cell Foundation Models: A Systematic Benchmark Across Architectures and Long-Tail Loss Functions

The paper benchmarks six long‑tail loss functions—cross‑entropy, weighted CE, class‑balanced loss, focal loss, LDAM, and logit‑adjusted softmax—across three single‑cell foundation model architectures (scGPT, scBERT, Geneformer) and three datasets (Multiple Sclerosis, Zheng68K, human Pancreas). It shows that overall accuracy masks systematic failures on rare, disease‑relevant cell types, with a consistent gap between overall accuracy, Macro‑F1, and rare‑class recall under plain cross‑entropy. The study identifies two distinct regimes of rare‑class failure, predicts reweighting efficacy by absolute training‑set size, and finds class‑balanced loss and LDAM to be the most reliable across all settings.

By Zeyu Dong, Jiahui Zhong
ragbenchmarks
More like this →
arXiv Machine Learning
Sep 22

Efficient K-generalizable Learned Search

arXiv:2603.06159v2 Announce Type: replace-cross Abstract: Learned top-K search improves the accuracy-latency trade-off of graph-based vector search, but existing methods are designed for a fixed K: s...

By Yifan Peng, Jiafei Fan, Xingda Wei, Sijie Shen, Rong Chen, Jianning Wang, Xiaojian Luo, Wenyuan Yu, Jingren Zhou, Haibo Chen
rag
More like this →