Nerva leverages sparse matrix operations via Intel MKL to accelerate neural network training, reducing time by 4× at 99% sparsity with comparable accuracy.

Wieger Wesselink, Bram Grooten, Qiao Xiao et al.

Introduced ebc-exstream, leveraging model explanations and human feedback to detect spurious correlations in concept drift, reducing annotation costs.

Cristiana Lalletti, Stefano Teso

SNIGL model leverages PNS theory to extract necessary and sufficient invariant subgraphs, boosting graph OOD generalization.

Xuexin Chen, Ruichu Cai, Kaitao Zheng et al.

Reinforcement learning (e.g., PPO, path consistency) fine-tunes diffusion models to maximize downstream rewards, applicable in bio-sequences and image generation.

Masatoshi Uehara, Yulai Zhao, Tommaso Biancalani et al.

STAD employs Bayesian filtering within state-space models to adapt to gradually evolving temporal distribution shifts, outperforming traditional TTA methods.

Mona Schirmer, Dan Zhang, Eric Nalisnick

ProtoFlow combines concept networks and generative classifiers, achieving 99.36% accuracy with interpretability and invertibility.

Zachariah Carmichael, Timothy Redgrave, Daniel Gonzalez Cedre et al.

Introduces non-Euclidean geometry, topology, and algebra into machine learning, enabling analysis of complex structured data with geometric, topological, and algebraic methods.

Mathilde Papillon, Sophia Sanborn, Johan Mathe et al.

EfficientQAT combines Block-AP and E2E-QP, significantly improving low-bit quantization of large models with reduced training time.

Mengzhao Chen, Wenqi Shao, Peng Xu et al.

Proposes 'Knowledge Boosting' to enhance small models during low-latency inference using delayed hints from remote large models, achieving up to 3.53dB SI-SDR gain.

Vidya Srinivas, Malek Itani, Tuochao Chen et al.

This paper introduces debate as a scalable oversight protocol, using weak LLMs to judge strong LLMs, showing superior performance in information-asymmetry tasks with up to 8% accuracy gain.

Zachary Kenton, Noah Y. Siegel, János Kramár et al.

NEBULA combines VQ-VAE and neural empirical Bayes for rapid, scalable 3D molecular library generation with strong generalization.

Ewa M. Nowara, Pedro O. Pinheiro, Sai Pooja Mahajan et al.

DisCo-Diff integrates discrete latent variables into diffusion models, reducing ODE curvature and improving image generation, achieving SOTA FID 1.65 on ImageNet-128.

Yilun Xu, Gabriele Corso, Tommi Jaakkola et al.

This study revisits classical NCA, integrating deep features and neighbor sampling, achieving top performance on 300 datasets comparable to CatBoost.

Han-Jia Ye, Huai-Hong Yin, De-Chuan Zhan et al.

Layer swapping and deletion experiments reveal four universal inference stages in LLMs, showing high robustness especially in middle layers, with detailed neural mechanisms identified.

Vedang Lad, Jin Hwa Lee, Wes Gurnee et al.

Proposed Stochastic Concept Bottleneck Model (SCBM) models concept dependencies via multivariate normal distribution, significantly improving intervention effectiveness on synthetic and real datasets.

Moritz Vandenhirtz, Sonia Laguna, Ričards Marcinkevičs et al.

RouteLLM trains preference-based routers, reducing costs over 2x while maintaining response quality.

Isaac Ong, Amjad Almahairi, Vincent Wu et al.

Univariate Skeleton Prediction using Transformers excels in multivariate systems, surpassing existing methods.

Giorgio Morales, John W. Sheppard

DriftLens is an unsupervised real-time concept drift detection framework with detection accuracy up to 85%.

Salvatore Greco, Bartolomeo Vacchetti, Daniele Apiletti et al.

This study introduces a similarity-based pruning method to remove 50% of Attention layers in Llama-2-70B, achieving 48.4% speedup with only 2.4% performance loss.

Shwai He, Guoheng Sun, Zheyu Shen et al.

LayerMerge jointly prunes and merges convolutional and activation layers, achieving 1.36× speed-up with minimal accuracy loss.

Jinuk Kim, Marwa El Halabi, Mingi Ji et al.