AI Papers
Browse and discover the latest research papers on artificial intelligence, machine learning, and related fields.
202
PaliGemma: A versatile 3B VLM for transfer
Lucas Beyer, Andreas Steiner, Andr'e Susano Pinto, Alexander Kolesnikov, Xiao Wang, Daniel Salz, Maxim Neumann, Ibrahim Alabdulmohsin, Michael Tschannen, Emanuele Bugliarello, Thomas Unterthiner, Daniel Keysers, Skanda Koppula, Fangyu Liu, Adam Grycner, Alexey Gritsenko, Neil Houlsby, Manoj Kumar, Keran Rong, Julian Eisenschlos, Rishabh Kabra, Matthias Bauer, Matko Bov{s}njak, Xi Chen, Matthias Minderer, Paul Voigtlaender, Ioana Bica, Ivana Balazevic, Joan Puigcerver, Pinelopi Papalampidi, Olivier Henaff, Xi Xiong, Radu Soricut, Jeremiah Harmsen, Xiaohua Zhai
PaliGemma is an open Vision-Language Model (VLM) that is based on the SigLIP-So400m vision encoder and the Gemma-2B language model. It is trained to be a versatile and broadly knowledgeable base model that is effective to transfer. It achieves strong performance on a wide variety of open-world tasks. We evaluate PaliGemma on almost 40 diverse tasks including standard VLM benchmarks, but also more specialized tasks such as remote-sensing and segmentation.
Read more10/11/2024
201
New!More Agents Is All You Need
Junyou Li, Qin Zhang, Yangbin Yu, Qiang Fu, Deheng Ye
We find that, simply via a sampling-and-voting method, the performance of large language models (LLMs) scales with the number of agents instantiated. Also, this method, termed as Agent Forest, is orthogonal to existing complicated methods to further enhance LLMs, while the degree of enhancement is correlated to the task difficulty. We conduct comprehensive experiments on a wide range of LLM benchmarks to verify the presence of our finding, and to study the properties that can facilitate its occurrence. Our code is publicly available at: https://github.com/MoreAgentsIsAllYouNeed/AgentForest
Read more10/15/2024
86
The Geometry of Categorical and Hierarchical Concepts in Large Language Models
Kiho Park, Yo Joong Choe, Yibo Jiang, Victor Veitch
The linear representation hypothesis is the informal idea that semantic concepts are encoded as linear directions in the representation spaces of large language models (LLMs). Previous work has shown how to make this notion precise for representing binary concepts that have natural contrasts (e.g., {male, female}) as directions in representation space. However, many natural concepts do not have natural contrasts (e.g., whether the output is about an animal). In this work, we show how to extend the formalization of the linear representation hypothesis to represent features (e.g., is_animal) as vectors. This allows us to immediately formalize the representation of categorical concepts as polytopes in the representation space. Further, we use the formalization to prove a relationship between the hierarchical structure of concepts and the geometry of their representations. We validate these theoretical results on the Gemma and LLaMA-3 large language models, estimating representations for 900+ hierarchically related concepts using data from WordNet.
Read more10/10/2024
67
Aria: An Open Multimodal Native Mixture-of-Experts Model
Dongxu Li, Yudong Liu, Haoning Wu, Yue Wang, Zhiqi Shen, Bowen Qu, Xinyao Niu, Guoyin Wang, Bei Chen, Junnan Li
Information comes in diverse modalities. Multimodal native AI models are essential to integrate real-world information and deliver comprehensive understanding. While proprietary multimodal native models exist, their lack of openness imposes obstacles for adoptions, let alone adaptations. To fill this gap, we introduce Aria, an open multimodal native model with best-in-class performance across a wide range of multimodal, language, and coding tasks. Aria is a mixture-of-expert model with 3.9B and 3.5B activated parameters per visual token and text token, respectively. It outperforms Pixtral-12B and Llama3.2-11B, and is competitive against the best proprietary models on various multimodal tasks. We pre-train Aria from scratch following a 4-stage pipeline, which progressively equips the model with strong capabilities in language understanding, multimodal understanding, long context window, and instruction following. We open-source the model weights along with a codebase that facilitates easy adoptions and adaptations of Aria in real-world applications.
Read more10/15/2024
🤿
64
New!Standalone 16-bit Training: Missing Study for Hardware-Limited Deep Learning Practitioners
Juyoung Yun, Sol Choi, Francois Rameau, Byungkon Kang, Zhoulai Fu
With the increasing complexity of machine learning models, managing computational resources like memory and processing power has become a critical concern. Mixed precision techniques, which leverage different numerical precisions during model training and inference to optimize resource usage, have been widely adopted. However, access to hardware that supports lower precision formats (e.g., FP8 or FP4) remains limited, especially for practitioners with hardware constraints. For many with limited resources, the available options are restricted to using 32-bit, 16-bit, or a combination of the two. While it is commonly believed that 16-bit precision can achieve results comparable to full (32-bit) precision, this study is the first to systematically validate this assumption through both rigorous theoretical analysis and extensive empirical evaluation. Our theoretical formalization of floating-point errors and classification tolerance provides new insights into the conditions under which 16-bit precision can approximate 32-bit results. This study fills a critical gap, proving for the first time that standalone 16-bit precision neural networks match 32-bit and mixed-precision in accuracy while boosting computational speed. Given the widespread availability of 16-bit across GPUs, these findings are especially valuable for machine learning practitioners with limited hardware resources to make informed decisions.
Read more10/15/2024
63
It's Your Turn: A Novel Channel Contention Mechanism for Improving Wi-Fi's Reliability
Francesc Wilhelmi, Lorenzo Galati-Giordano, Gianluca Fontanesi
The next generation of Wi-Fi, i.e., the IEEE 802.11bn (aka Wi-Fi 8), is not only expected to increase its performance and provide extended capabilities but also aims to offer a reliable service. Given that one of the main sources of unreliability in IEEE 802.11 stems from the current distributed channel access, which is based on Listen-Before-Talk (LBT), the development of novel contention schemes gains importance for Wi-Fi 8 and beyond. In this paper, we propose a new channel contention mechanism, It's Your Turn (IYT), that extends the existing Distributed Coordination Function (DCF) and aims at improving the reliability of distributed LBT by providing ordered device transmissions thanks to neighboring activity awareness. Using simulation results, we show that our mechanism strives to provide reliable performance by controlling the channel access delay. We prove the versatility of IYT against different topologies, coexistence with legacy devices, and increasing network densities.
Read more10/11/2024
44
New!Long Context Compression with Activation Beacon
Peitian Zhang, Zheng Liu, Shitao Xiao, Ninglu Shao, Qiwei Ye, Zhicheng Dou
Long context compression is a critical research problem due to its significance in reducing the high computational and memory costs associated with LLMs. In this paper, we propose Activation Beacon, a plug-in module for transformer-based LLMs that targets effective, efficient, and flexible compression of long contexts. To achieve this, our method introduces the following technical designs. 1) We directly compress the activations (i.e. keys and values at every layer), rather than leveraging soft prompts to relay information (which constitute a major bottleneck to encapsulate the complex information within long contexts). 2) We tailor the compression workflow, where each fine-grained input unit is progressively compressed, enabling high-quality compression and efficient computation during both training and inference. 3) We train the model through compression-based auto-regression, making full use of plain texts and instructional data to optimize the model's compression performance. 4) During training, we randomly sample a compression ratio at each step, teaching the model to support a wide range of compression configurations. Extensive evaluations are conducted on various long-context tasks whose lengths (e.g., 128K) may far exceed the maximum training length (20K), such as document understanding, few-shot learning, and Needle-in-a-Haystack. Whilst existing methods struggle to handle these challenging tasks, Activation Beacon maintains a comparable performance to the uncompressed baseline across various scenarios, achieving a 2x acceleration in inference time and an 8x reduction of memory costs for KV cache. Our data, model, and code have been released at url{https://github.com/FlagOpen/FlagEmbedding/}.
Read more10/15/2024
📉
43
New!Meissonic: Revitalizing Masked Generative Transformers for Efficient High-Resolution Text-to-Image Synthesis
Jinbin Bai, Tian Ye, Wei Chow, Enxin Song, Qing-Guo Chen, Xiangtai Li, Zhen Dong, Lei Zhu, Shuicheng Yan
Diffusion models, such as Stable Diffusion, have made significant strides in visual generation, yet their paradigm remains fundamentally different from autoregressive language models, complicating the development of unified language-vision models. Recent efforts like LlamaGen have attempted autoregressive image generation using discrete VQVAE tokens, but the large number of tokens involved renders this approach inefficient and slow. In this work, we present Meissonic, which elevates non-autoregressive masked image modeling (MIM) text-to-image to a level comparable with state-of-the-art diffusion models like SDXL. By incorporating a comprehensive suite of architectural innovations, advanced positional encoding strategies, and optimized sampling conditions, Meissonic substantially improves MIM's performance and efficiency. Additionally, we leverage high-quality training data, integrate micro-conditions informed by human preference scores, and employ feature compression layers to further enhance image fidelity and resolution. Our model not only matches but often exceeds the performance of existing models like SDXL in generating high-quality, high-resolution images. Extensive experiments validate Meissonic's capabilities, demonstrating its potential as a new standard in text-to-image synthesis. We release a model checkpoint capable of producing $1024 times 1024$ resolution images.
Read more10/15/2024
41
Smaller, Weaker, Yet Better: Training LLM Reasoners via Compute-Optimal Sampling
Hritik Bansal, Arian Hosseini, Rishabh Agarwal, Vinh Q. Tran, Mehran Kazemi
Training on high-quality synthetic data from strong language models (LMs) is a common strategy to improve the reasoning performance of LMs. In this work, we revisit whether this strategy is compute-optimal under a fixed inference budget (e.g., FLOPs). To do so, we investigate the trade-offs between generating synthetic data using a stronger but more expensive (SE) model versus a weaker but cheaper (WC) model. We evaluate the generated data across three key metrics: coverage, diversity, and false positive rate, and show that the data from WC models may have higher coverage and diversity, but also exhibit higher false positive rates. We then finetune LMs on data from SE and WC models in different settings: knowledge distillation, self-improvement, and a novel weak-to-strong improvement setup where a weaker LM teaches reasoning to a stronger LM. Our findings reveal that models finetuned on WC-generated data consistently outperform those trained on SE-generated data across multiple benchmarks and multiple choices of WC and SE models. These results challenge the prevailing practice of relying on SE models for synthetic data generation, suggesting that WC may be the compute-optimal approach for training advanced LM reasoners.
Read more10/10/2024
39
Skip Hash: A Fast Ordered Map Via Software Transactional Memory
Matthew Rodriguez, Vitaly Aksenov, Michael Spear
Scalable ordered maps must ensure that range queries, which operate over many consecutive keys, provide intuitive semantics (e.g., linearizability) without degrading the performance of concurrent insertions and removals. These goals are difficult to achieve simultaneously when concurrent data structures are built using only locks and compare-and-swap objects. However, recent innovations in software transactional memory (STM) allow programmers to assume that multi-word atomic operations can be fast and simple. This paper introduces the skip hash, a new ordered map designed around that assumption. It combines a skip list and a hash map behind a single abstraction, resulting in $O(1)$ overheads for most operations. The skip hash makes use of a novel range query manager -- again leveraging STM -- to achieve fast, linearizable range queries that do not inhibit scalability. In performance evaluation, we show that the skip hash outperforms the state of the art in almost all cases. This places the skip hash in the uncommon position of being both exceedingly fast and exceedingly simple.
Read more10/11/2024
âž–
38
MOMENT: A Family of Open Time-series Foundation Models
Mononito Goswami, Konrad Szafer, Arjun Choudhry, Yifu Cai, Shuo Li, Artur Dubrawski
We introduce MOMENT, a family of open-source foundation models for general-purpose time series analysis. Pre-training large models on time series data is challenging due to (1) the absence of a large and cohesive public time series repository, and (2) diverse time series characteristics which make multi-dataset training onerous. Additionally, (3) experimental benchmarks to evaluate these models, especially in scenarios with limited resources, time, and supervision, are still in their nascent stages. To address these challenges, we compile a large and diverse collection of public time series, called the Time series Pile, and systematically tackle time series-specific challenges to unlock large-scale multi-dataset pre-training. Finally, we build on recent work to design a benchmark to evaluate time series foundation models on diverse tasks and datasets in limited supervision settings. Experiments on this benchmark demonstrate the effectiveness of our pre-trained models with minimal data and task-specific fine-tuning. Finally, we present several interesting empirical observations about large pre-trained time series models. Pre-trained models (AutonLab/MOMENT-1-large) and Time Series Pile (AutonLab/Timeseries-PILE) are available on Huggingface.
Read more10/11/2024
📈
16
New!Uncovering mesa-optimization algorithms in Transformers
Johannes von Oswald, Maximilian Schlegel, Alexander Meulemans, Seijin Kobayashi, Eyvind Niklasson, Nicolas Zucchet, Nino Scherrer, Nolan Miller, Mark Sandler, Blaise Aguera y Arcas, Max Vladymyrov, Razvan Pascanu, Jo~ao Sacramento
Some autoregressive models exhibit in-context learning capabilities: being able to learn as an input sequence is processed, without undergoing any parameter changes, and without being explicitly trained to do so. The origins of this phenomenon are still poorly understood. Here we analyze a series of Transformer models trained to perform synthetic sequence prediction tasks, and discover that standard next-token prediction error minimization gives rise to a subsidiary learning algorithm that adjusts the model as new inputs are revealed. We show that this process corresponds to gradient-based optimization of a principled objective function, which leads to strong generalization performance on unseen sequences. Our findings explain in-context learning as a product of autoregressive loss minimization and inform the design of new optimization-based Transformer layers.
Read more10/16/2024