20 papers · ranked by Valyu relevance
Wang, Zhichao, Ma, Dongyang + 14 more
The "end-to-end" label for LLMs is a misnomer. In practice, they depend on a nondifferentiable decoding process that requires laborious, hand-tuning of hyperparameters like temperature and top-p. This paper introduces AutoDeco, a novel architecture that enables truly "end-to-end" generation by learning to control its…
Yuhang Wang, Weihua Chen, Linjing Song, Zhiping Xu + 6 more
With the rapid growth of data volume in sensor networks, lossy source coding systems achieve high-efficiency data compression with low distortion under limited transmission bandwidth. However, conventional compression algorithms rely on a two-stage framework with high computational complexity and frequently struggle to…
Yongbin Mu, Mieradilijiang Maimaiti, Miaomiao Xu, Wenkai Li + 2 more
Scene text recognition has significant application value in autonomous driving, smart retail, and assistive devices. However, due to challenges such as multi-scale variations, distortions, and complex backgrounds, existing methods such as CRNN, ViT, and PARSeq, while showing good performance, still have room for…
Ahmed El-Gazzar, Marcel van Gerven
The rapid growth of large-scale neuroscience datasets has spurred diverse modeling strategies, ranging from mechanistic models grounded in biophysics, to phenomenological descriptions of neural dynamics, to data-driven deep neural networks (DNNs). Each approach offers distinct strengths as mechanistic models provide…
Stéphane d’Ascoli, Corentin Bel, Jérémy Rapin, Hubert Banville + 3 more
While deep learning has enabled the decoding of language from intracranial brain recordings, achieving this with non-invasive recordings remains an open challenge. We introduce a deep learning pipeline to decode individual words from electro- (EEG) and magneto-encephalography (MEG) signals. We evaluate our approach on…
Ori Hendler, Ronen Segev, Maoz Shamir
Sensory information propagates through successive processing stages in the brain, where synaptic weight patterns between stations determine how downstream neurons decode information from upstream populations. Although optimized synaptic connectivity can enhance information transmission, it requires precise weight…
G. Bilodeau, A. Miao, G. Gagnon-Turcotte, C. Ethier + 1 more
Bidirectional interfaces combined with neural de-coding algorithms are essential for closed-loop (CL) neuromodulation, enabling simultaneous neural monitoring and responsive optogenetic stimulation. However, implementing these capabilities in compact wireless headstages for freely moving animals remains challenging, as…
Authors not listed
The discovery of chemically novel or structurally anomalous metal-organic frameworks (MOFs) is essential for expanding reticular design space and enhancing dataset reliability. We present CHEM-AD (Chemically Unusual Metal–organic Frameworks via Autoencoder-based Detection), a label-free, CPU-efficient pipeline that…
Alon Helvits, Eliya Nachmani
Error-correcting codes enable reliable communication, yet practical soft decoding remains challenging across code families and block lengths. We propose SB-ECC, a score-based decoder that casts decoding as continuous-time denoising. A neural denoiser defines a probability-flow ordinary differential equation (ODE) that…
Ashena Gorgan Mohammadi, Manu Srinath Halvagal, Friedemann Zenke
Tracking prey or recognizing a lurking predator is as crucial for survival as anticipating their actions. To guide behavior, the brain must extract information about object identities and their dynamics from entangled sensory inputs. How it accomplishes this feat remains an open question. Predictive coding theories…
Raphaël Le Bidan, Ahmad Ismail, Elsa Dupraz, Charbel Abdel Nour
Syndrome-based neural decoding (SBND) has emerged as a promising deep learning approach for soft-decision decoding of high-rate, short-length codes. However, this approach still has substantial room for improvement. In this paper, we show how to leverage code automorphisms to enhance the ability of existing SBND models…
Zhenkai Qin, Baozhong Wei, Caifeng Gao, Jianyuan Ni + 1 more
Time series forecasting is essential in energy, finance, and meteorology. However, existing Transformer-based models face challenges with computational inefficiency and poor generalization for long-term sequences. To address these issues, this study proposes the KEDformer framework. It integrates knowledge extraction…
Francisco Miqueles, Adrián G. Palacios, John Atkinson, María-José Escobar
Introduction Understanding how deep learning models map neural population activity to stimuli requires both high predictive accuracy and interpretable internal mechanisms. Methods In this work, we employ the POYO framework, a scalable transformer architecture based on spike tokenization and latent modeling, to decode…
Aviv Adler, Jennifer Tang
It is well-known in the field of lossless data compression that probabilistic nextsymbol prediction can be used to compress sequences of symbols. Deep neural networks are able to capture rich dependencies in data, offering a powerful means of estimating these probabilities and hence an avenue towards more effective…
Shi Luohe, Li, Zuchao, Zhang + 5 more
Speculative decoding accelerates LLM inference by utilizing otherwise idle computational resources during memory-tochip data transfer. Current speculative decoding methods typically assume a considerable amount of available computing power, then generate a complex and massive draft tree using a small autoregressive…
Changsoo Shin
Modern AI systems excel at pattern recognition and task execution, but they often fall short of replicating the layered, self-referential structure of human thought that unfolds over time. In this paper, we present a mathematically grounded and conceptually simple framework based on smoothed step functions-sigmoid…
Authors not listed
Large Language Models have demonstrated impressive capabilities in natural language understanding and processing. However, as AI and LLMs continue to evolve, their ability to accurately and efficiently interpret data from scientific figures and plots remains obscure. In this study, we test and evaluate the ability of…
Eliezer Masliah
How transient neural representations become integrated and stable enough to function as internal neural models remains incompletely understood. Grounded in efficient coding, Bayesian and predictive frameworks, recurrent and attractor dynamics, neural state-space models, and systems neuroscience, the Principle of…
Vasanth Kumar Babu, Arno Granier, Timothée Proix, Marmaduke Woodman + 4 more
The neocortex is central to mammalian cognition, yet a computational framework that is both biologically constrained and capable of performing complex cognitive tasks remains missing. Here we show that cortico-thalamic circuits are well suited to implement multi-head self- and cross-attention, the mechanism underlying…
Authors not listed
Meta-GGA density functional theory (DFT) is an important method in ab initio materials modelling; however, its computational cost limits applicability for generating large datasets or simulating extended length and time scales, as necessary for modern materials discovery. Deorbitalization is a promising strategy to…