Tootfinder

Spain training Ukrainian military personnel on servicing Patriot missile system: https://benborges.xyz/2024/02/02/spain-training-ukrainian.html

How the EFF, Techdirt, MuckRock, and DDoSecrets are pushing back against legal threats aiming to censor reports on Appin's alleged hacker-for-hire past (Andy Greenberg/Wired)
https://www.wired.com/story/appin-training-centers-lawsuits-censorship/…

A Startup Allegedly ‘Hacked the World.’ Then Came the Censorship—and Now the Backlash
A loose coalition of anti-censorship voices is working to highlight reports of one Indian company’s hacker-for-hire past—and the legal threats aimed at making them disappear.

Inhalte von Wordpress.com und Tumblr sollen für KI-Training freigegeben werden
Auch Automattic will mit nutzergenerierten Inhalten Geld machen und sie fürs KI-Training freigeben. Wer das nicht will, muss tätig werden.

Fine Structure-Aware Sampling: A New Sampling Training Scheme for Pixel-Aligned Implicit Models in Single-View Human Reconstruction
Kennard Yanting Chan, Fayao Liu, Guosheng Lin, Chuan Sheng Foo, Weisi Lin
https://arxiv.org/abs/2402.19197

Fine Structure-Aware Sampling: A New Sampling Training Scheme for Pixel-Aligned Implicit Models in Single-View Human Reconstruction
Pixel-aligned implicit models, such as PIFu, PIFuHD, and ICON, are used for single-view clothed human reconstruction. These models need to be trained using a sampling training scheme. Existing sampling training schemes either fail to capture thin surfaces (e.g. ears, fingers) or cause noisy artefacts in reconstructed meshes. To address these problems, we introduce Fine Structured-Aware Sampling (FSS), a new sampling training scheme to train pixel-aligned implicit models for single-view human re…

Wiley licenses content for training an #LLM. The company was not named, but I would suspect it's the one which has been signing a lot of licensing deals lately. Access to STM content could be a big differentiator, though I wouldn't expect it to be exclusive. Also, $23M sounds small.

Q3 2024 John Wiley & Sons Inc Earnings Call
Q3 2024 John Wiley & Sons Inc Earnings Call

This https://arxiv.org/abs/2403.13799 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csCL_…

Reverse Training to Nurse the Reversal Curse
Large language models (LLMs) have a surprising failure: when trained on "A has a feature B", they do not generalize to "B is a feature of A", which is termed the Reversal Curse. Even when training with trillions of tokens this issue still appears due to Zipf's law - hence even if we train on the entire internet. This work proposes an alternative training scheme, called reverse training, whereby all words are used twice, doubling the amount of available tokens. The LLM is trained in both forward…

Sources: OpenAI discussed training GPT-5 on public YouTube video transcripts; AI companies struggle to find quality training data as publishers block access (Deepa Seetharaman/Wall Street Journal)
https://www.…

Updated Page: Study and Training https://www.philippawrites.co.uk/home/study-and-training

Study and Training - Philippa Writes
The ongoing study and training that Philippa Willitts undertakes to ensure that she is on top of her specialist subjects and has a wide breadth of other knowledge.

Navigating WebAI: Training Agents to Complete Web Tasks with Large Language Models and Reinforcement Learning
Lucas-Andre\"i Thil, Mirela Popa, Gerasimos Spanakis
https://arxiv.org/abs/2405.00516

Navigating WebAI: Training Agents to Complete Web Tasks with Large Language Models and Reinforcement Learning
Recent advancements in language models have demonstrated remarkable improvements in various natural language processing (NLP) tasks such as web navigation. Supervised learning (SL) approaches have achieved impressive performance while utilizing significantly less training data compared to previous methods. However, these SL-based models fall short when compared to reinforcement learning (RL) approaches, which have shown superior results. In this paper, we propose a novel approach that combines …

🔊 #NowPlaying on BBCRadio2's #ClaudiaWinkleman
Dua Lipa:
🎵 Training Season
#DuaLipa
https://open.spotify.com/track/2UrKwpJwcoOlYZpooWD1hY
https://sielrecords.bandcamp.com/track/dua-lipa-training-season-2

A pretest-posttest pilot study for augmented reality-based physical-cognitive training in community-dwelling older adults at risk of mild cognitive impairment
Sirinun Chaipunko, Watthanaree Ammawat, Keerathi Oanmun, Wanvipha Hongnaphadol, Supatida Sorasak, Pattrawadee Makmee
https://arxiv.org/abs/2404.18970 https://arxiv.org/pdf/2404.18970
arXiv:2404.18970v1 Announce Type: new
Abstract: As cognitive interventions for older adults evolve, modern technologies are increasingly integrated into their development. This study investigates the efficacy of augmented reality (AR)-based physical-cognitive training using an interactive game with Kinect motion sensor technology on older individuals at risk of mild cognitive impairment. Utilizing a pretest-posttest experimental design, twenty participants (mean age 66.8 SD. = 4.6 years, age range 60-78 years) underwent eighteen individual training sessions, lasting 45 to 60 minutes each, conducted three times a week over a span of 1.5 months. The training modules from five activities, encompassing episodic and working memory, attention and inhibition, cognitive flexibility, and speed processing, were integrated with physical movement and culturally relevant Thai-context activities. Results revealed significant improvements in inhibition, cognitive flexibility, accuracy, and reaction time, with working memory demonstrating enhancements in accuracy albeit not in reaction time. These findings underscore the potential of AR interventions to bolster basic executive enhancement among community-dwelling older adults at risk of cognitive decline.

Lancet: Accelerating Mixture-of-Experts Training via Whole Graph Computation-Communication Overlapping
Chenyu Jiang, Ye Tian, Zhen Jia, Shuai Zheng, Chuan Wu, Yida Wang
https://arxiv.org/abs/2404.19429

Lancet: Accelerating Mixture-of-Experts Training via Whole Graph Computation-Communication Overlapping
The Mixture-of-Expert (MoE) technique plays a crucial role in expanding the size of DNN model parameters. However, it faces the challenge of extended all-to-all communication latency during the training process. Existing methods attempt to mitigate this issue by overlapping all-to-all with expert computation. Yet, these methods frequently fall short of achieving sufficient overlap, consequently restricting the potential for performance enhancements. In our study, we extend the scope of this cha…

Always be Pre-Training: Representation Learning for Network Intrusion Detection with GNNs
Zhengyao Gu, Diego Troy Lopez, Lilas Alrahis, Ozgur Sinanoglu
https://arxiv.org/abs/2402.18986

Always be Pre-Training: Representation Learning for Network Intrusion Detection with GNNs
Graph neural network-based network intrusion detection systems have recently demonstrated state-of-the-art performance on benchmark datasets. Nevertheless, these methods suffer from a reliance on target encoding for data pre-processing, limiting widespread adoption due to the associated need for annotated labels--a cost-prohibitive requirement. In this work, we propose a solution involving in-context pre-training and the utilization of dense representations for categorical features to jointly o…

Update from Ukraine | Great news! Ruzzian Training base and Oil refinery Kapputed. ATACMS in action: https://benborges.xyz/2024/05/02/update-from-ukraine.html

I’m giving a (fairly exploratory) talk on the LLM training and inference stack and how to monitor it at Multicore World, Christchurch New Zealand on Feb 14th. We’re getting there and back by island hopping to explore some new places. Arrived in Tahiti today.

Yes, Your Sports Bra Really Can Restrict Your Breathing
https://www.outsideonline.com/health/training-performance/sports-bra-breathing-restriction-study/

Not new and not limited to cyber. I have seen it for decades. Do more with less money, fewer bodies and little to no training. It appears to work until ripe organic bovine waste product hits the fan and then the ones at the bottom get screwed over and fired.
https://www.…

Closed-loop training of static output feedback neural network controllers for large systems: A distillation case study
E. M. Turan, J. J\"aschke
https://arxiv.org/abs/2402.19309

Closed-loop training of static output feedback neural network controllers for large systems: A distillation case study
The online implementation of model predictive control for constrained multivariate systems has two main disadvantages: it requires an estimate of the entire model state and an optimisation problem must be solved online. These issues have typically been treated separately. This work proposes an integrated approach for the offline training of an output feedback neural network controller in closed loop. Online this neural network controller computers the plant inputs cheaply using noisy measuremen…

I just let Claude write a short article about :has() as a test and while it got browser support wrong (probably because of older training data) the code examples turned out far better than those in the latest piece by DigitalOcean on CSS-Tricks… 🫣

I am not a fan of having everything I create on every platform which I don’t completely own being auto-opt-in for them to monetize my labor for LLM (fake-#AI) training.
https://front-end.social/@benschwarz/1

benschwarz (@benschwarz@front-end.social)
Attached: 1 image Huggingface "BigCode" AI project did a shitty thing where they indexed public code from GitHub, making it an OPT-OUT experience. You can see if your account is included and request removal (along with 1,800+ other disappointed souls) here: https://huggingface.co/spaces/bigcode/in-the-stack I had 81 repositories included, and while my more important work is private, I am still sending eye-daggers towards huggingface's general direction. You can do it too! Or just opt-out,…

Dual Dynamic Threshold Adjustment Strategy for Deep Metric Learning
Xiruo Jiang, Yazhou Yao, Sheng Liu, Fumin Shen, Liqiang Nie, Xiansheng Hua
https://arxiv.org/abs/2404.19282 https://arxiv.org/pdf/2404.19282
arXiv:2404.19282v1 Announce Type: new
Abstract: Loss functions and sample mining strategies are essential components in deep metric learning algorithms. However, the existing loss function or mining strategy often necessitate the incorporation of additional hyperparameters, notably the threshold, which defines whether the sample pair is informative. The threshold provides a stable numerical standard for determining whether to retain the pairs. It is a vital parameter to reduce the redundant sample pairs participating in training. Nonetheless, finding the optimal threshold can be a time-consuming endeavor, often requiring extensive grid searches. Because the threshold cannot be dynamically adjusted in the training stage, we should conduct plenty of repeated experiments to determine the threshold. Therefore, we introduce a novel approach for adjusting the thresholds associated with both the loss function and the sample mining strategy. We design a static Asymmetric Sample Mining Strategy (ASMS) and its dynamic version Adaptive Tolerance ASMS (AT-ASMS), tailored for sample mining methods. ASMS utilizes differentiated thresholds to address the problems (too few positive pairs and too many redundant negative pairs) caused by only applying a single threshold to filter samples. AT-ASMS can adaptively regulate the ratio of positive and negative pairs during training according to the ratio of the currently mined positive and negative pairs. This meta-learning-based threshold generation algorithm utilizes a single-step gradient descent to obtain new thresholds. We combine these two threshold adjustment algorithms to form the Dual Dynamic Threshold Adjustment Strategy (DDTAS). Experimental results show that our algorithm achieves competitive performance on CUB200, Cars196, and SOP datasets.

Jahrelanges Training im Münchener Nahverkehr und Bahnhöfen hat es mir ermöglicht den optimalen Umstieg mit kurze Laufweg so hinzukriegen um den, für den DB Navigator unmöglichen, Umstieg hinzukriegen. Grüße aus dem ICE 802 😁

If you liked our last year's short paper on #DeepLearning from #TrajectoryData, you'll love our new #preprint even more:
📝 "MobilityDL: A Review of Deep Learning From Trajecto…

MobilityDL: A Review of Deep Learning From Trajectory Data
Trajectory data combines the complexities of time series, spatial data, and (sometimes irrational) movement behavior. As data availability and computing power have increased, so has the popularity of deep learning from trajectory data. This review paper provides the first comprehensive overview of deep learning approaches for trajectory data. We have identified eight specific mobility use cases which we analyze with regards to the deep learning models and the training data used. Besides a compr…

Fake it to make it: Using synthetic data to remedy the data shortage in joint multimodal speech-and-gesture synthesis
Shivam Mehta, Anna Deichler, Jim O'Regan, Birger Mo\"ell, Jonas Beskow, Gustav Eje Henter, Simon Alexanderson
https://arxiv.org/abs/2404.19622 https://arxiv.org/pdf/2404.19622
arXiv:2404.19622v1 Announce Type: new
Abstract: Although humans engaged in face-to-face conversation simultaneously communicate both verbally and non-verbally, methods for joint and unified synthesis of speech audio and co-speech 3D gesture motion from text are a new and emerging field. These technologies hold great promise for more human-like, efficient, expressive, and robust synthetic communication, but are currently held back by the lack of suitably large datasets, as existing methods are trained on parallel data from all constituent modalities. Inspired by student-teacher methods, we propose a straightforward solution to the data shortage, by simply synthesising additional training material. Specifically, we use unimodal synthesis models trained on large datasets to create multimodal (but synthetic) parallel training data, and then pre-train a joint synthesis model on that material. In addition, we propose a new synthesis architecture that adds better and more controllable prosody modelling to the state-of-the-art method in the field. Our results confirm that pre-training on large amounts of synthetic data improves the quality of both the speech and the motion synthesised by the multimodal model, with the proposed architecture yielding further benefits when pre-trained on the synthetic data. See https://shivammehta25.github.io/MAGI/ for example output.

Automatic Cardiac Pathology Recognition in Echocardiography Images Using Higher Order Dynamic Mode Decomposition and a Vision Transformer for Small Datasets
Andr\'es Bell-Navas, Nourelhouda Groun, Mar\'ia Villalba-Orero, Enrique Lara-Pezzi, Jes\'us Garicano-Mena, Soledad Le Clainche
https://arxiv.org/abs/2404.19579 https://arxiv.org/pdf/2404.19579
arXiv:2404.19579v1 Announce Type: new
Abstract: Heart diseases are the main international cause of human defunction. According to the WHO, nearly 18 million people decease each year because of heart diseases. Also considering the increase of medical data, much pressure is put on the health industry to develop systems for early and accurate heart disease recognition. In this work, an automatic cardiac pathology recognition system based on a novel deep learning framework is proposed, which analyses in real-time echocardiography video sequences. The system works in two stages. The first one transforms the data included in a database of echocardiography sequences into a machine-learning-compatible collection of annotated images which can be used in the training stage of any kind of machine learning-based framework, and more specifically with deep learning. This includes the use of the Higher Order Dynamic Mode Decomposition (HODMD) algorithm, for the first time to the authors' knowledge, for both data augmentation and feature extraction in the medical field. The second stage is focused on building and training a Vision Transformer (ViT), barely explored in the related literature. The ViT is adapted for an effective training from scratch, even with small datasets. The designed neural network analyses images from an echocardiography sequence to predict the heart state. The results obtained show the superiority of the proposed system and the efficacy of the HODMD algorithm, even outperforming pretrained Convolutional Neural Networks (CNNs), which are so far the method of choice in the literature.

This https://arxiv.org/abs/2402.08523 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_qu…

Investigating the Effect of Noise on the Training Performance of Hybrid Quantum Neural Networks
In this paper, we conduct a comprehensively analyze the influence of different quantum noise gates, including Phase Flip, Bit Flip, Phase Damping, Amplitude Damping, and the Depolarizing Channel, on the performance of HyQNNs. Our results reveal distinct and significant effects on HyQNNs training and validation accuracies across different probabilities of noise. For instance, the Phase Flip gate introduces phase errors, and we observe that HyQNNs exhibit resilience at higher probability (p = 1.0…

Das #Kundenglücklichmachungsposteinlieferungsgeräusch findet zum Glück auch alleine raus...
Frau #Atelierhütehund frönt dem Training für #faulsterborderderwelt.

Saints to hold 2024 training camp at University of California, Irvine https://www.nfl.com/news/saints-to-hold-2024-training-camp-at-university-of-california-irvine

Saints to hold 2024 training camp at University of California, Irvine
The New Orleans Saints are moving training camp to the University of California, Irvine, in 2024, the team announced Friday.

Navigating WebAI: Training Agents to Complete Web Tasks with Large Language Models and Reinforcement Learning
Lucas-Andre\"i Thil, Mirela Popa, Gerasimos Spanakis
https://arxiv.org/abs/2405.00516

Navigating WebAI: Training Agents to Complete Web Tasks with Large Language Models and Reinforcement Learning
Recent advancements in language models have demonstrated remarkable improvements in various natural language processing (NLP) tasks such as web navigation. Supervised learning (SL) approaches have achieved impressive performance while utilizing significantly less training data compared to previous methods. However, these SL-based models fall short when compared to reinforcement learning (RL) approaches, which have shown superior results. In this paper, we propose a novel approach that combines …

This https://arxiv.org/abs/2308.14947 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csRO_…

Improving Generalization in Reinforcement Learning Training Regimes for Social Robot Navigation
In order for autonomous mobile robots to navigate in human spaces, they must abide by our social norms. Reinforcement learning (RL) has emerged as an effective method to train sequential decision-making policies that are able to respect these norms. However, a large portion of existing work in the field conducts both RL training and testing in simplistic environments. This limits the generalization potential of these models to unseen environments, and the meaningfulness of their reported result…

This makes me happy. I've always like Dom. I was at his very first spring training game for the Mets.
MLB Insider: Red Sox to sign former Mets, Cubs first baseman Dom Smith https://fansided.com/posts/mlb-insider-red-so…

MLB Insider: Red Sox to sign former Mets, Cubs first baseman Dom Smith
Free-agent first baseman Dom Smith and the Boston Red Sox are in agreement on a major-league contract, according to sources familiar with the deal.

I’m attending The Democratic National Committee’s event, “All Aboard: the Training Series - What's at stake with the extreme MAGA GOP agenda” – sign up now to join me! https://events.democrats.org/event/607046/?

All Aboard: the Training Series - What's at stake with the extreme MAGA GOP agenda · The Democratic National Committee
As Joe and Kamala say, it's time to “finish the job!” All aboard! We are thrilled to invite you to the last of a three part training series on messaging and having effective conversations on the phones, at the doors, and with our friends and family about the upcoming election. This third session will focus on how to have persuasive conversations with voters that show what's at stake with the extreme MAGA GOP agenda this year. Please RSVP to receive Zoom details. **Check out the other…

This Microsoft Research paper seems to indicate that you can pretrain an LLM such that it is 2.71 times faster and uses 3.55 times less GPU memory. Also saves 71.4 times arithmetic operations energy consumption.
#ai #energy #gpu

"KI befindet sich auf einem technischen Plateau, #gpt5 wird nicht merklich besser sein als GPT-4, wenn es überhaupt jemals erscheint"
>> Es passiert nicht oft, dass man eine "Alchemie-Phase" einer Technologie live miterleben kann, wo die Anwendung die erklärende Kraft der dahinterliegenden Theorien (zeitweise) übersteigt.
Momentan werden immer noch wöchentlic…

The Era of 1-bit LLMs: All Large Language Models are in 1.58 Bits
Recent research, such as BitNet, is paving the way for a new era of 1-bit Large Language Models (LLMs). In this work, we introduce a 1-bit LLM variant, namely BitNet b1.58, in which every single parameter (or weight) of the LLM is ternary {-1, 0, 1}. It matches the full-precision (i.e., FP16 or BF16) Transformer LLM with the same model size and training tokens in terms of both perplexity and end-task performance, while being significantly more cost-effective in terms of latency, memory, through…

EyeEm: Fotos löschen, sonst KI-Training
Wer nicht möchte, dass seine Fotos zu Trainingsmaterial für KI werden, muss sie bei der Plattform EyeEm löschen. Die AGB sehen das so vor.
https://www.h…

Socrates said:
“No man has the right to be an amateur in the matter of physical training. It is a shame for a man to grow old without seeing the beauty and strength of which his body is capable.”
The Greeks invented #calisthenics, the term coming from the words kšllos (κάλλος), "beauty," and sthenos (σθένος), "strength." It is the art of using one's body …

Joint Training and Reflection Pattern Optimization for Non-Ideal RIS-Aided Multiuser Systems
Zhenyao He, Jindan Xu, Hong Shen, Wei Xu, Chau Yuen, Marco Di Renzo
https://arxiv.org/abs/2403.19955

Joint Training and Reflection Pattern Optimization for Non-Ideal RIS-Aided Multiuser Systems
Reconfigurable intelligent surface (RIS) is a promising technique to improve the performance of future wireless communication systems at low energy consumption. To reap the potential benefits of RIS-aided beamforming, it is vital to enhance the accuracy of channel estimation. In this paper, we consider an RIS-aided multiuser system with non-ideal reflecting elements, each of which has a phase-dependent reflecting amplitude, and we aim to minimize the mean-squared error (MSE) of the channel esti…

#OH on slack:
#infosec #cybersecurity

Better & Faster Large Language Models via Multi-token Prediction
Fabian Gloeckle, Badr Youbi Idrissi, Baptiste Rozi\`ere, David Lopez-Paz, Gabriel Synnaeve
https://arxiv.org/abs/2404.19737 https://arxiv.org/pdf/2404.19737
arXiv:2404.19737v1 Announce Type: new
Abstract: Large language models such as GPT and Llama are trained with a next-token prediction loss. In this work, we suggest that training language models to predict multiple future tokens at once results in higher sample efficiency. More specifically, at each position in the training corpus, we ask the model to predict the following n tokens using n independent output heads, operating on top of a shared model trunk. Considering multi-token prediction as an auxiliary training task, we measure improved downstream capabilities with no overhead in training time for both code and natural language models. The method is increasingly useful for larger model sizes, and keeps its appeal when training for multiple epochs. Gains are especially pronounced on generative benchmarks like coding, where our models consistently outperform strong baselines by several percentage points. Our 13B parameter models solves 12 % more problems on HumanEval and 17 % more on MBPP than comparable next-token models. Experiments on small algorithmic tasks demonstrate that multi-token prediction is favorable for the development of induction heads and algorithmic reasoning capabilities. As an additional benefit, models trained with 4-token prediction are up to 3 times faster at inference, even with large batch sizes.

Registration is open for the online course “Ministering to Those Impacted by Suicide.” The course meets 4 Mondays in a row, beginning April 15.
It is estimated that 50% of individuals in any congregation have been impacted by suicide in some way. This training will take a deep dive into the topic and expand the thinking of leaders about how churches and organizations can proactively address the hopelessness and isolation that contribute to suicidality.

One model to use them all: Training a segmentation model with complementary datasets
Alexander C. Jenke, Sebastian Bodenstedt, Fiona R. Kolbinger, Marius Distler, J\"urgen Weitz, Stefanie Speidel
https://arxiv.org/abs/2402.19340

One model to use them all: Training a segmentation model with complementary datasets
Understanding a surgical scene is crucial for computer-assisted surgery systems to provide any intelligent assistance functionality. One way of achieving this scene understanding is via scene segmentation, where every pixel of a frame is classified and therefore identifies the visible structures and tissues. Progress on fully segmenting surgical scenes has been made using machine learning. However, such models require large amounts of annotated training data, containing examples of all relevant…

Scaling and renormalization in high-dimensional regression
Alexander B. Atanasov, Jacob A. Zavatone-Veth, Cengiz Pehlevan
https://arxiv.org/abs/2405.00592 …

Scaling and renormalization in high-dimensional regression
This paper presents a succinct derivation of the training and generalization performance of a variety of high-dimensional ridge regression models using the basic tools of random matrix theory and free probability. We provide an introduction and review of recent results on these topics, aimed at readers with backgrounds in physics and deep learning. Analytic formulas for the training and generalization errors are obtained in a few lines of algebra directly from the properties of the $S$-transfor…

Active Learning with Task Adaptation Pre-training for Speech Emotion Recognition
Dongyuan Li, Ying Zhang, Yusong Wang, Funakoshi Kataro, Manabu Okumura
https://arxiv.org/abs/2405.00307

Active Learning with Task Adaptation Pre-training for Speech Emotion Recognition
Speech emotion recognition (SER) has garnered increasing attention due to its wide range of applications in various fields, including human-machine interaction, virtual assistants, and mental health assistance. However, existing SER methods often overlook the information gap between the pre-training speech recognition task and the downstream SER task, resulting in sub-optimal performance. Moreover, current methods require much time for fine-tuning on each specific speech dataset, such as IEMOCA…

📢 Ver.di ruft Lufthansa-Bodenpersonal zu mehrtägigem Warnstreik auf
Im Tarifkonflikt um das Lufthansa-Bodenpersonal hat die Gewerkschaft ver.di einen mehrtägigen Warnstreik ab Mittwoch angekündigt. Zum dreitägigen bundesweiten Ausstand bis Freitag aufgerufen seien unter anderem Beschäftigte und Auszubildende der Lufthansa Technik, Lufthansa Aviation Training und Lufthansa Technical Training, teilte ver.di mit.
➡️

Ver.di ruft Lufthansa-Bodenpersonal zu mehrtägigem Warnstreik auf
Im Tarifkonflikt um das Lufthansa-Bodenpersonal hat die Gewerkschaft ver.di einen mehrtägigen Warnstreik ab Mittwoch angekündigt. Zum dreitägigen bundesweiten Ausstand bis Freitag aufgerufen seien unter anderem Beschäftigte und Auszubildende der Lufthansa Technik, Lufthansa Aviation Training und Lufthansa Technical Training, teilte ver.di mit.

Employing Federated Learning for Training Autonomous HVAC Systems
Fredrik Hagstr\"om, Vikas Garg, Fabricio Oliveira
https://arxiv.org/abs/2405.00389 h…

Employing Federated Learning for Training Autonomous HVAC Systems
Buildings account for 40 % of global energy consumption. A considerable portion of building energy consumption stems from heating, ventilation, and air conditioning (HVAC), and thus implementing smart, energy-efficient HVAC systems has the potential to significantly impact the course of climate change. In recent years, model-free reinforcement learning algorithms have been increasingly assessed for this purpose due to their ability to learn and adapt purely from experience. They have been shown…

“Isometric exercises can help reduce blood pressure even more effectively than aerobic activity, weight training, or high-intensity interval workouts.”
#Health #Fitness #Longevity

The surprisingly simple exercise that can lower your blood pressure
The wall sit, an isometric exercise, has been shown to help lower blood pressure better than other types of exercise.

Sources: OpenAI discussed training GPT-5 on public YouTube video transcripts; AI industry's need for high-quality text data may outstrip supply within two years (Deepa Seetharaman/Wall Street Journal)
https://www.

I will give a research talk today (Friday, the 1st of March) at 11:00 (CET = UTC 1) in our local Functional Programming research group seminar.
There is a Zoom link for the stream which should hopefully work: https://chalmers.zoom.us/j/65586341322 (Password: f,p.t:a;l,k without the pun…

Early call to gauge interest: Who's seriously interested in attending a paid training on oscilloscope probing theory and practice, some time in the early summer at my lab just outside Seattle? It will be a one-day, in-person event including 4 hours of lecture and lots of hands-on lab time for everyone on some very nice equipment (16 GHz oscilloscope, 28 Gbps BERT, multiple VNAs, and more).
I've done a few test runs with friends and I think I've finally got it refined to th…

electronics-training/oscilloscope-probing at master · azonenberg/electronics-training
Training materials for classes etc. Contribute to azonenberg/electronics-training development by creating an account on GitHub.

EEG classifier cross-task transfer to avoid training sessions in robot-assisted rehabilitation
Niklas Kueper, Su Kyoung Kim, Elsa Andrea Kirchner
https://arxiv.org/abs/2402.17790 …

EEG classifier cross-task transfer to avoid training sessions in robot-assisted rehabilitation
Background: For an individualized support of patients during rehabilitation, learning of individual machine learning models from the human electroencephalogram (EEG) is required. Our approach allows labeled training data to be recorded without the need for a specific training session. For this, the planned exoskeleton-assisted rehabilitation enables bilateral mirror therapy, in which movement intentions can be inferred from the activity of the unaffected arm. During this therapy, labeled EEG da…

Leak Proof CMap; a framework for training and evaluation of cell line agnostic L1000 similarity methods
Steven Shave, Richard Kasprowicz, Abdullah M. Athar, Denise Vlachou, Neil O. Carragher, Cuong Q. Nguyen
https://arxiv.org/abs/2404.18960

Leak Proof CMap; a framework for training and evaluation of cell line agnostic L1000 similarity methods
The Connectivity Map (CMap) is a large publicly available database of cellular transcriptomic responses to chemical and genetic perturbations built using a standardized acquisition protocol known as the L1000 technique. Databases such as CMap provide an exciting opportunity to enrich drug discovery efforts, providing a 'known' phenotypic landscape to explore and enabling the development of state of the art techniques for enhanced information extraction and better informed decisions. Whilst mult…

I'm delighted to be launching 'Mastodon for organisations' training on behalf of @…
https://afallen.cymru/2024/04/24/intro

Spyx: A Library for Just-In-Time Compiled Optimization of Spiking Neural Networks
Kade M. Heckel, Thomas Nowotny
https://arxiv.org/abs/2402.18994 https://<…

Spyx: A Library for Just-In-Time Compiled Optimization of Spiking Neural Networks
As the role of artificial intelligence becomes increasingly pivotal in modern society, the efficient training and deployment of deep neural networks have emerged as critical areas of focus. Recent advancements in attention-based large neural architectures have spurred the development of AI accelerators, facilitating the training of extensive, multi-billion parameter models. Despite their effectiveness, these powerful networks often incur high execution costs in production environments. Neuromor…

"Training Data for the Price of a Sandwich – Common Crawl’s Impact on Generative AI "
https://foundation.mozilla.org/de/research/library/generative-ai-training-data/common-crawl/

Training Data for the Price of a Sandwich: Common Crawl’s Impact on Generative AI
Mozilla research finds that Common Crawl's outsized role in the generative AI boom has improved transparency and competition, but is also contributing to biased and opaque generative AI models.

Can DC Reporters Overcome Their Trumper Shock-Training? https://talkingpointsmemo.com/edblog/can-dc-reporters-overcome-their-trumper-shock-training

Can DC Reporters Overcome Their Trumper Shock-Training?
As I noted yesterday, the Alexander Smirnov news is either confirmation of what we already knew or else spurs a kind of mass outburst of incredulity about how it is we're still as a country…

Alterstherapie und Rehazentren setzen auf spielerische, Videospiel-basierte Trainings,
#Demenz #Training

Exergaming: Was Videospiel-basiertes Fitnesstraining bringt
Übungen an Spielkonsolen helfen Körper und Geist. Rehakliniken und Schweizer Start-ups können erste Erfolge vorweisen.

Had to sign into facebook for first time in *years* for a kid's thing.
Man that site is *terrible* 🤢
Even worse than instagram.
Overloaded by adverts, "celebrity" stories and "AI" spam pretending to be real photos... I can only hope these count as training critical thinking.. But for goodness sake, WHY do people still insist on using it...?

Training Dynamics of Multi-Head Softmax Attention for In-Context Learning: Emergence, Convergence, and Optimality
Siyu Chen, Heejune Sheen, Tianhao Wang, Zhuoran Yang
https://arxiv.org/abs/2402.19442

Training Dynamics of Multi-Head Softmax Attention for In-Context Learning: Emergence, Convergence, and Optimality
We study the dynamics of gradient flow for training a multi-head softmax attention model for in-context learning of multi-task linear regression. We establish the global convergence of gradient flow under suitable choices of initialization. In addition, we prove that an interesting "task allocation" phenomenon emerges during the gradient flow dynamics, where each attention head focuses on solving a single task of the multi-task model. Specifically, we prove that the gradient flow dynamics can b…

I just let Claude write a short article about :has() as a test and while it got browser support wrong (probably because of older training data) the code examples turned out far better than those in the latest piece by DigitalOcean on CSS-Tricks… 🫣

This https://arxiv.org/abs/2401.17010 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csCR_…

Finetuning Large Language Models for Vulnerability Detection
This paper presents the results of finetuning large language models (LLMs) for the task of detecting vulnerabilities in source code. We leverage WizardCoder, a recent improvement of the state-of-the-art LLM StarCoder, and adapt it for vulnerability detection through further finetuning. To accelerate training, we modify WizardCoder's training procedure, also we investigate optimal training regimes. For the imbalanced dataset with many more negative examples than positive, we also explore differe…

Towards Tracing Trustworthiness Dynamics: Revisiting Pre-training Period of Large Language Models
Chen Qian, Jie Zhang, Wei Yao, Dongrui Liu, Zhenfei Yin, Yu Qiao, Yong Liu, Jing Shao
https://arxiv.org/abs/2402.19465

Towards Tracing Trustworthiness Dynamics: Revisiting Pre-training Period of Large Language Models
Ensuring the trustworthiness of large language models (LLMs) is crucial. Most studies concentrate on fully pre-trained LLMs to better understand and improve LLMs' trustworthiness. In this paper, to reveal the untapped potential of pre-training, we pioneer the exploration of LLMs' trustworthiness during this period, focusing on five key dimensions: reliability, privacy, toxicity, fairness, and robustness. To begin with, we apply linear probing to LLMs. The high probing accuracy suggests that \te…

If you liked our last year's short paper on #DeepLearning from #TrajectoryData, you'll love our new #preprint even more:
📝 "MobilityDL: A Review of Deep Learning From Trajecto…

MobilityDL: A Review of Deep Learning From Trajectory Data
Trajectory data combines the complexities of time series, spatial data, and (sometimes irrational) movement behavior. As data availability and computing power have increased, so has the popularity of deep learning from trajectory data. This review paper provides the first comprehensive overview of deep learning approaches for trajectory data. We have identified eight specific mobility use cases which we analyze with regards to the deep learning models and the training data used. Besides a compr…

Elliptic, MIT, and IBM release an experimental new AI detection model and its 200M-transaction training dataset to help identify Bitcoin money laundering (Andy Greenberg/Wired)
https://www.wired.com/story/ai-crypto-tracing-model-money-laundering/

This https://arxiv.org/abs/2310.09617 has been replaced.
link: https://scholar.google.com/scholar?q=a

How Good is ChatGPT in Giving Advice on Your Visualization Design?
Data visualization practitioners often lack formal training, resulting in a knowledge gap in visualization design best practices. Large-language models like ChatGPT, with their vast internet-scale training data, offer transformative potential in addressing this gap. To explore this potential, we adopted a mixed-method approach. Initially, we analyzed the VisGuide forum, a repository of data visualization questions, by comparing ChatGPT-generated responses to human replies. Subsequently, our use…

This makes me happy. I've always like Dom. I was at his very first spring training game for the Mets.
MLB Insider: Red Sox to sign former Mets, Cubs first baseman Dom Smith https://fansided.com/posts/mlb-insider-red-so…

MLB Insider: Red Sox to sign former Mets, Cubs first baseman Dom Smith
Free-agent first baseman Dom Smith and the Boston Red Sox are in agreement on a major-league contract, according to sources familiar with the deal.

Training Generative Image Super-Resolution Models by Wavelet-Domain Losses Enables Better Control of Artifacts
Cansu Korkmaz, A. Murat Tekalp, Zafer Dogan
https://arxiv.org/abs/2402.19215

Training Generative Image Super-Resolution Models by Wavelet-Domain Losses Enables Better Control of Artifacts
Super-resolution (SR) is an ill-posed inverse problem, where the size of the set of feasible solutions that are consistent with a given low-resolution image is very large. Many algorithms have been proposed to find a "good" solution among the feasible solutions that strike a balance between fidelity and perceptual quality. Unfortunately, all known methods generate artifacts and hallucinations while trying to reconstruct high-frequency (HF) image details. A fundamental question is: Can a model l…

This https://arxiv.org/abs/2404.12538 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csCV_…

TrACT: A Training Dynamics Aware Contrastive Learning Framework for Long-tail Trajectory Prediction
As a safety critical task, autonomous driving requires accurate predictions of road users' future trajectories for safe motion planning, particularly under challenging conditions. Yet, many recent deep learning methods suffer from a degraded performance on the challenging scenarios, mainly because these scenarios appear less frequently in the training data. To address such a long-tail issue, existing methods force challenging scenarios closer together in the feature space during training to tri…

Moldovan military holds JCET drills with US, Romania: https://benborges.xyz/2024/04/01/moldovan-military-holds.html

Saints moving camp due to renovation of 'F' caf https://www.espn.com/nfl/story/_/id/39833555/saints-moving-training-camp-due-renovation-f-cafeteria

Saints moving training camp to renovate F-graded cafeteria - ESPN
The New Orleans Saints announced they are moving their 2024 training camp to the University of California, Irvine due to construction of their new cafeteria.

This https://arxiv.org/abs/2310.13828 has been replaced.
link: https://scholar.google.com/scholar?q=a

Nightshade: Prompt-Specific Poisoning Attacks on Text-to-Image Generative Models
Data poisoning attacks manipulate training data to introduce unexpected behaviors into machine learning models at training time. For text-to-image generative models with massive training datasets, current understanding of poisoning attacks suggests that a successful attack would require injecting millions of poison samples into their training pipeline. In this paper, we show that poisoning attacks can be successful on generative models. We observe that training data per concept can be quite lim…

I'm delighted to be launching 'Mastodon for organisations' training on behalf of @…
https://afallen.cymru/2024/04/24/intro

A look at Project Maven, the US DOD's flagship AI effort which identifies battlefield targets, and at concerns, including adversaries poisoning training data (Katrina Manson/Bloomberg)
https://www.bloomberg.com/features/2024-ai-warfare-project-maven/…

AI Warfare Becomes Real for US Military With Project Maven
US military operators started out skeptical about AI, but now they are the ones developing and using Project Maven to identify targets on the battlefield.

ArCHer: Training Language Model Agents via Hierarchical Multi-Turn RL
Yifei Zhou, Andrea Zanette, Jiayi Pan, Sergey Levine, Aviral Kumar
https://arxiv.org/abs/2402.19446

ArCHer: Training Language Model Agents via Hierarchical Multi-Turn RL
A broad use case of large language models (LLMs) is in goal-directed decision-making tasks (or "agent" tasks), where an LLM needs to not just generate completions for a given prompt, but rather make intelligent decisions over a multi-turn interaction to accomplish a task (e.g., when interacting with the web, using tools, or providing customer support). Reinforcement learning (RL) provides a general paradigm to address such agent tasks, but current RL methods for LLMs largely focus on optimizing…

Had to sign into facebook for first time in *years* for a kid's thing.
Man that site is *terrible* 🤢
Even worse than instagram.
Overloaded by adverts, "celebrity" stories and "AI" spam pretending to be real photos... I can only hope these count as training critical thinking.. But for goodness sake, WHY do people still insist on using it...?

Training Classical Neural Networks by Quantum Machine Learning
Chen-Yu Liu, En-Jui Kuo, Chu-Hsuan Abraham Lin, Sean Chen, Jason Gemsun Young, Yeong-Jar Chang, Min-Hsiu Hsieh
https://arxiv.org/abs/2402.16465

Training Classical Neural Networks by Quantum Machine Learning
In recent years, advanced deep neural networks have required a large number of parameters for training. Therefore, finding a method to reduce the number of parameters has become crucial for achieving efficient training. This work proposes a training scheme for classical neural networks (NNs) that utilizes the exponentially large Hilbert space of a quantum system. By mapping a classical NN with $M$ parameters to a quantum neural network (QNN) with $O(\text{polylog} (M))$ rotational gate angles, …

Safe Training with Sensitive In-domain Data: Leveraging Data Fragmentation To Mitigate Linkage Attacks
Mariia Ignashina, Julia Ive
https://arxiv.org/abs/2404.19486 https://arxiv.org/pdf/2404.19486
arXiv:2404.19486v1 Announce Type: new
Abstract: Current text generation models are trained using real data which can potentially contain sensitive information, such as confidential patient information and the like. Under certain conditions output of the training data which they have memorised can be triggered, exposing sensitive data. To mitigate against this risk we propose a safer alternative which sees fragmented data in the form of domain-specific short phrases randomly grouped together shared instead of full texts. Thus, text fragments that could re-identify an individual cannot be reproduced by the model in one sequence, giving significant protection against linkage attacks. We fine-tune several state-of-the-art LLMs using meaningful syntactic chunks to explore their utility. In particular, we fine-tune BERT-based models to predict two cardiovascular diagnoses. Our results demonstrate the capacity of LLMs to benefit from the pre-trained knowledge and deliver classification results when fine-tuned with fragmented data comparable to fine-tuning with full training data.

Leveraging AI Predicted and Expert Revised Annotations in Interactive Segmentation: Continual Tuning or Full Training?
Tiezheng Zhang, Xiaoxi Chen, Chongyu Qu, Alan Yuille, Zongwei Zhou
https://arxiv.org/abs/2402.19423

Leveraging AI Predicted and Expert Revised Annotations in Interactive Segmentation: Continual Tuning or Full Training?
Interactive segmentation, an integration of AI algorithms and human expertise, premises to improve the accuracy and efficiency of curating large-scale, detailed-annotated datasets in healthcare. Human experts revise the annotations predicted by AI, and in turn, AI improves its predictions by learning from these revised annotations. This interactive process continues to enhance the quality of annotations until no major revision is needed from experts. The key challenge is how to leverage AI pred…

This https://arxiv.org/abs/2304.01705 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_ees…

Cross-modal tumor segmentation using generative blending augmentation and self training
\textit{Objectives}: Data scarcity and domain shifts lead to biased training sets that do not accurately represent deployment conditions. A related practical problem is cross-modal image segmentation, where the objective is to segment unlabelled images using previously labelled datasets from other imaging modalities. \textit{Methods}: We propose a cross-modal segmentation method based on conventional image synthesis boosted by a new data augmentation technique called Generative Blending Augment…

Homomorphic WiSARDs: Efficient Weightless Neural Network training over encrypted data
Leonardo Neumann, Antonio Guimar\~aes, Diego F. Aranha, Edson Borin
https://arxiv.org/abs/2403.20190

Homomorphic WiSARDs: Efficient Weightless Neural Network training over encrypted data
The widespread application of machine learning algorithms is a matter of increasing concern for the data privacy research community, and many have sought to develop privacy-preserving techniques for it. Among existing approaches, the homomorphic evaluation of ML algorithms stands out by performing operations directly over encrypted data, enabling strong guarantees of confidentiality. The homomorphic evaluation of inference algorithms is practical even for relatively deep Convolution Neural Netw…

This https://arxiv.org/abs/2404.10255 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csLG_…

Privacy-Preserving Training-as-a-Service for On-Device Intelligence: Concept, Architectural Scheme, and Open Problems
On-device intelligence (ODI) enables artificial intelligence (AI) applications to run on end devices, providing real-time and customized AI services without relying on remote servers. However, training models for on-device deployment face significant challenges due to the decentralized and privacy-sensitive nature of users' data, along with end-side constraints related to network connectivity, computation efficiency, etc. Existing training paradigms, such as cloud-based training, federated lear…

This https://arxiv.org/abs/2311.08590 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csCL_…

PEMA: An Offsite-Tunable Plug-in External Memory Adaptation for Language Models
Pre-trained language models (PLMs) show impressive performance in various downstream NLP tasks. However, pre-training large language models demands substantial memory and training compute. Furthermore, due to the substantial resources required, many PLM weights are confidential. Consequently, users are compelled to share their data with model owners for fine-tuning specific tasks. To overcome the limitations, we introduce Plug-in External Memory Adaptation (PEMA), a Parameter-Efficient Fine-Tun…

MTLoRA: A Low-Rank Adaptation Approach for Efficient Multi-Task Learning
Ahmed Agiza, Marina Neseem, Sherief Reda
https://arxiv.org/abs/2403.20320 https://…

MTLoRA: A Low-Rank Adaptation Approach for Efficient Multi-Task Learning
Adapting models pre-trained on large-scale datasets to a variety of downstream tasks is a common strategy in deep learning. Consequently, parameter-efficient fine-tuning methods have emerged as a promising way to adapt pre-trained models to different tasks while training only a minimal number of parameters. While most of these methods are designed for single-task adaptation, parameter-efficient training in Multi-Task Learning (MTL) architectures is still unexplored. In this paper, we introduce …

KI-Training mit Büchern: OpenAI mit Teilerfolg bei Copyright-Klage von Autoren
Ein Gericht weist einige Punkte einer Autoren-Klage wegen Nutzung von Büchern für KI-Training ab. Der Vorwurf der direkten Urheberrechtsverletzung aber bleibt.

Training Classical Neural Networks by Quantum Machine Learning
Chen-Yu Liu, En-Jui Kuo, Chu-Hsuan Abraham Lin, Sean Chen, Jason Gemsun Young, Yeong-Jar Chang, Min-Hsiu Hsieh
https://arxiv.org/abs/2402.16465

Training Classical Neural Networks by Quantum Machine Learning
In recent years, advanced deep neural networks have required a large number of parameters for training. Therefore, finding a method to reduce the number of parameters has become crucial for achieving efficient training. This work proposes a training scheme for classical neural networks (NNs) that utilizes the exponentially large Hilbert space of a quantum system. By mapping a classical NN with $M$ parameters to a quantum neural network (QNN) with $O(\text{polylog} (M))$ rotational gate angles, …

This https://arxiv.org/abs/2304.01705 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_ees…

Cross-modal tumor segmentation using generative blending augmentation and self training
\textit{Objectives}: Data scarcity and domain shifts lead to biased training sets that do not accurately represent deployment conditions. A related practical problem is cross-modal image segmentation, where the objective is to segment unlabelled images using previously labelled datasets from other imaging modalities. \textit{Methods}: We propose a cross-modal segmentation method based on conventional image synthesis boosted by a new data augmentation technique called Generative Blending Augment…

Extending Llama-3's Context Ten-Fold Overnight
Peitian Zhang, Ninglu Shao, Zheng Liu, Shitao Xiao, Hongjin Qian, Qiwei Ye, Zhicheng Dou
https://arxiv.org/abs/2404.19553 https://arxiv.org/pdf/2404.19553
arXiv:2404.19553v1 Announce Type: new
Abstract: We extend the context length of Llama-3-8B-Instruct from 8K to 80K via QLoRA fine-tuning. The entire training cycle is super efficient, which takes 8 hours on one 8xA800 (80G) GPU machine. The resulted model exhibits superior performances across a broad range of evaluation tasks, such as NIHS, topic retrieval, and long-context language understanding; meanwhile, it also well preserves the original capability over short contexts. The dramatic context extension is mainly attributed to merely 3.5K synthetic training samples generated by GPT-4 , which indicates the LLMs' inherent (yet largely underestimated) potential to extend its original context length. In fact, the context length could be extended far beyond 80K with more computation resources. Therefore, the team will publicly release the entire resources (including data, model, data generation pipeline, training code) so as to facilitate the future research from the community: \url{https://github.com/FlagOpen/FlagEmbedding}.

VIXEN: Visual Text Comparison Network for Image Difference Captioning
Alexander Black, Jing Shi, Yifei Fai, Tu Bui, John Collomosse
https://arxiv.org/abs/2402.19119

VIXEN: Visual Text Comparison Network for Image Difference Captioning
We present VIXEN - a technique that succinctly summarizes in text the visual differences between a pair of images in order to highlight any content manipulation present. Our proposed network linearly maps image features in a pairwise manner, constructing a soft prompt for a pretrained large language model. We address the challenge of low volume of training data and lack of manipulation variety in existing image difference captioning (IDC) datasets by training on synthetically manipulated images…

This https://arxiv.org/abs/2402.00786 has been replaced.
link: https://scholar.google.com/scholar?q=a

CroissantLLM: A Truly Bilingual French-English Language Model
We introduce CroissantLLM, a 1.3B language model pretrained on a set of 3T English and French tokens, to bring to the research and industrial community a high-performance, fully open-sourced bilingual model that runs swiftly on consumer-grade local hardware. To that end, we pioneer the approach of training an intrinsically bilingual model with a 1:1 English-to-French pretraining data ratio, a custom tokenizer, and bilingual finetuning datasets. We release the training dataset, notably containin…

Distributed Stochastic Optimization of a Neural Representation Network for Time-Space Tomography Reconstruction
K. Aditya Mohan, Massimiliano Ferrucci, Chuck Divin, Garrett A. Stevenson, Hyojin Kim
https://arxiv.org/abs/2404.19075 https://arxiv.org/pdf/2404.19075
arXiv:2404.19075v1 Announce Type: new
Abstract: 4D time-space reconstruction of dynamic events or deforming objects using X-ray computed tomography (CT) is an extremely ill-posed inverse problem. Existing approaches assume that the object remains static for the duration of several tens or hundreds of X-ray projection measurement images (reconstruction of consecutive limited-angle CT scans). However, this is an unrealistic assumption for many in-situ experiments that causes spurious artifacts and inaccurate morphological reconstructions of the object. To solve this problem, we propose to perform a 4D time-space reconstruction using a distributed implicit neural representation (DINR) network that is trained using a novel distributed stochastic training algorithm. Our DINR network learns to reconstruct the object at its output by iterative optimization of its network parameters such that the measured projection images best match the output of the CT forward measurement model. We use a continuous time and space forward measurement model that is a function of the DINR outputs at a sparsely sampled set of continuous valued object coordinates. Unlike existing state-of-the-art neural representation architectures that forward and back propagate through dense voxel grids that sample the object's entire time-space coordinates, we only propagate through the DINR at a small subset of object coordinates in each iteration resulting in an order-of-magnitude reduction in memory and compute for training. DINR leverages distributed computation across several compute nodes and GPUs to produce high-fidelity 4D time-space reconstructions even for extremely large CT data sizes. We use both simulated parallel-beam and experimental cone-beam X-ray CT datasets to demonstrate the superior performance of our approach.

Alibaba is cutting prices for over 100 core cloud products in China by up to 55%, and 20% on average, including data storage and elastic computing products (Bloomberg)
https://www.bloomberg.com/news/articles/2024-02…

Alibaba Unveils Big Cloud Price Cuts as AI Rivalry Deepens
Alibaba Group Holding Ltd. is rolling out its second major price cut for cloud services in years, aiming to win back users from rivals like Tencent Holdings Ltd. competing to provide the tools critical to training artificial intelligence.

This https://arxiv.org/abs/2402.17532 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csCL_…

Retrieval is Accurate Generation
Standard language models generate text by selecting tokens from a fixed, finite, and standalone vocabulary. We introduce a novel method that selects context-aware phrases from a collection of supporting documents. One of the most significant challenges for this paradigm shift is determining the training oracles, because a string of text can be segmented in various ways and each segment can be retrieved from numerous possible documents. To address this, we propose to initialize the training orac…

Alibaba is cutting prices for over 100 core cloud products in China by up to 55%, and 20% on average, including data storage and elastic computing products (Bloomberg)
https://www.bloomberg.com/news/articles/2024-02…

Alibaba Unveils Big Cloud Price Cuts as AI Rivalry Deepens
Alibaba Group Holding Ltd. is rolling out its second major price cut for cloud services in years, aiming to win back users from rivals like Tencent Holdings Ltd. competing to provide the tools critical to training artificial intelligence.

Challenges in Pre-Training Graph Neural Networks for Context-Based Fake News Detection: An Evaluation of Current Strategies and Resource Limitations
Gregor Donabauer, Udo Kruschwitz
https://arxiv.org/abs/2402.18179

Challenges in Pre-Training Graph Neural Networks for Context-Based Fake News Detection: An Evaluation of Current Strategies and Resource Limitations
Pre-training of neural networks has recently revolutionized the field of Natural Language Processing (NLP) and has before demonstrated its effectiveness in computer vision. At the same time, advances around the detection of fake news were mainly driven by the context-based paradigm, where different types of signals (e.g. from social media) form graph-like structures that hold contextual information apart from the news article to classify. We propose to merge these two developments by applying p…

Unsupervised Information Refinement Training of Large Language Models for Retrieval-Augmented Generation
Shicheng Xu, Liang Pang, Mo Yu, Fandong Meng, Huawei Shen, Xueqi Cheng, Jie Zhou
https://arxiv.org/abs/2402.18150

Unsupervised Information Refinement Training of Large Language Models for Retrieval-Augmented Generation
Retrieval-augmented generation (RAG) enhances large language models (LLMs) by incorporating additional information from retrieval. However, studies have shown that LLMs still face challenges in effectively using the retrieved information, even ignoring it or being misled by it. The key reason is that the training of LLMs does not clearly make LLMs learn how to utilize input retrieved texts with varied quality. In this paper, we propose a novel perspective that considers the role of LLMs in RAG …

S\~onajaht: Definition Embeddings and Semantic Search for Reverse Dictionary Creation
Aleksei Dorkin, Kairit Sirts
https://arxiv.org/abs/2404.19430 https://arxiv.org/pdf/2404.19430
arXiv:2404.19430v1 Announce Type: new
Abstract: We present an information retrieval based reverse dictionary system using modern pre-trained language models and approximate nearest neighbors search algorithms. The proposed approach is applied to an existing Estonian language lexicon resource, S\~onaveeb (word web), with the purpose of enhancing and enriching it by introducing cross-lingual reverse dictionary functionality powered by semantic search.
The performance of the system is evaluated using both an existing labeled English dataset of words and definitions that is extended to contain also Estonian and Russian translations, and a novel unlabeled evaluation approach that extracts the evaluation data from the lexicon resource itself using synonymy relations.
Evaluation results indicate that the information retrieval based semantic search approach without any model training is feasible, producing median rank of 1 in the monolingual setting and median rank of 2 in the cross-lingual setting using the unlabeled evaluation approach, with models trained for cross-lingual retrieval and including Estonian in their training data showing superior performance in our particular task.