United States

Large Language Models (LLMs) have revolutionized natural language processing tasks. However, their practical application is constrained by substantial memory and computational demands. Post-training quantization (PTQ) is considered an effective method to accelerate LLM inference. Despite its growing popularity in LLM model compression, PTQ deployment faces two major challenges. First, low-bit quantization leads to performance degradation. Second, restricted by the limited integer computing unit type on GPUs, quantized matrix operations with different precisions cannot be effectively accelerated. To address these issues, we introduce a novel arbitrary-bit quantization algorithm and inference framework, ABQ-LLM. It achieves superior performance across various quantization settings and enables efficient arbitrary-precision quantized inference on the GPU. ABQ-LLM introduces several key innovations: (1) a distribution correction method for transformer blocks to mitigate distribution differences caused by full quantization of weights and activations, improving performance at low bit-widths. (2) the bit balance strategy to counteract performance degradation from asymmetric distribution issues at very low bit-widths (e.g., 2-bit). (3) an innovative quantization acceleration framework that reconstructs the quantization matrix multiplication of arbitrary precision combinations based on BTC (Binary TensorCore) equivalents, gets rid of the limitations of INT4/INT8 computing units. ABQ-LLM can convert each component bit width gain into actual acceleration gain, maximizing performance under mixed precision(e.g., W6A6, W2A8). Based on W2*A8 quantization configuration on LLaMA-7B model, it achieved a WikiText2 perplexity of 7.59 (2.17$\downarrow $ vs 9.76 in AffineQuant). Compared to SmoothQuant, we realized 1.6$\times$ acceleration improvement and 2.7$\times$ memory compression gain.

AAAI 2025

ABQ-LLM: Arbitrary-Bit Quantized Inference Acceleration for Large Language Models

learning on the edge

model compression

We are pleased to announce the Thirty-Eighth AAAI Conference on Artificial Intelligence (AAAI-24), which will be held in Vancouver, British Columbia at the Vancouver Convention Centre – West Building from 20-27 February, 2024.

The purpose of the AAAI conference series is to promote research in Artificial Intelligence (AI) and foster scientific exchange between researchers, practitioners, scientists, students, and engineers across the entirety of AI and its affiliated disciplines. AAAI-24 will feature technical paper presentations, special tracks, invited speakers, workshops, tutorials, poster sessions, senior member presentations, competitions, and exhibit programs, and a range of other activities to be announced.

We expect for AAAI-24 to be an in-person conference – one author of all accepted papers will be expected to present work in person unless there are exceptional circumstances that prevent this.

In order to access the AAAI-24 event page you need to register [here](https://aaai.org/aaai-conference/registration/)

AAAI 2024

The purpose of the AAAI conference series is to promote research in Artificial Intelligence (AI) and foster scientific exchange between researchers, practitioners, scientists, students, and engineers across the entirety of AI and its affiliated disciplines. 

poster

We are pleased to announce the Thirty-Ninth AAAI Conference on Artificial Intelligence (AAAI-25), which will be held in Philadelphia, Pennsylvania at the Pennsylvania Convention Center from February 25 to March 4, 2025.

The purpose of the AAAI conference series is to promote research in Artificial Intelligence (AI) and foster scientific exchange between researchers, practitioners, scientists, students, and engineers across the entirety of AI and its affiliated disciplines. AAAI-25 will feature technical paper presentations, special tracks, invited speakers, workshops, tutorials, poster sessions, senior member presentations, competitions, and exhibit programs, and a range of other activities to be announced.

### [Invited Speakers](https://aaai.org/conference/aaai/aaai-25/aaai-25-invited-speakers/)

Register [here](https://aaai.org/conference/aaai/aaai-25/registration/)

The purpose of the AAAI conference series is to promote research in Artificial Intelligence (AI) and foster scientific exchange between researchers, practitioners, scientists, students, and engineers across the entirety of AI and its affiliated disciplines. AAAI-25 will feature technical paper presentations, special tracks, invited speakers, workshops, tutorials, poster sessions, senior member presentations, competitions, and exhibit programs, and a range of other activities to be announced.



The powerful capability of HyperGraph Neural Networks (HGNNs) in modeling intricate, high-order relationships among multiple data samples stems primarily from their ability to aggregate both the direct neighborhood features of individual nodes and those associated with hyperedges. 
However, the limited scope of feature propagation in existing HGNNs significantly reduces the utilization of hypergraph information, exacerbating over-squashing and over-smoothing issues.
 To this end, we propose a novel $\boldsymbol{K}$-hop $\boldsymbol{H}$yper$\boldsymbol{G}$raph $\boldsymbol{N}$eural $\boldsymbol{N}$etwork (KHGNN) to facilitate the interactions of distant nodes and hyperedges. 
Specifically, the bisection nested convolution based on HyperGINE is employed to extract features from nodes, hyperedges, and structures along all shortest paths between nodes or hyperedges, providing representations of long-distance relationships. 
With these comprehensive path features, nodes and hyperedges are guided to aggregate distant information while learning their complex relationships. 
The extensive experiments, particularly on long-range graph datasets, demonstrate that the proposed method achieves SOTA performance compared to existing HGNNs and graph neural networks.

K-hop Hypergraph Neural Network: A Comprehensive Aggregation Approach

Multimodal large language models have experienced rapid growth, and numerous different models have emerged. The interpretability of MLLM remains an under-explored area. Especially when faced with more complex tasks such as chain-of-thought reasoning, its internal mechanisms still resemble a black box that is difficult to decipher. By studying the interaction and information flow between images and text, we noticed that in models such as LLaVA1.5,  image tokens that are semantically related to text are more likely to have information flow convergence in the LLM decoding layer, and these image tokens receive higher attention scores. However, those image tokens that are less relevant to the text do not have information flow convergence, and they only get very small attention scores. To efficiently utilize the image information, we propose a new image token reduction method, Simignore, which aims to improve the complex reasoning ability of MLLM by computing the similarity between image and text embeddings and ignoring image tokens that are irrelevant and unimportant to the text. Through extensive experiments, we demonstrate the effectiveness of our method for complex reasoning tasks.

Enhancing Multimodal Large Language Models Complex Reason via Similarity Computation

RAW-to-sRGB mapping, or the simulation of the traditional camera image signal processor (ISP), aims to generate DSLR-quality sRGB images from raw data captured by smartphone sensors. Despite achieving comparable results to sophisticated handcrafted camera ISP solutions, existing learning-based methods still struggle with detail disparity and color distortion. In this paper, we present ISPDiffuser, a diffusion-based decoupled framework that separates the RAW-to-sRGB mapping into detail reconstruction in grayscale space and color consistency mapping from grayscale to sRGB. Specifically, we propose a texture-aware diffusion model that leverages the generative ability of diffusion models to focus on local detail recovery, in which a texture enrichment loss is further proposed to prompt the diffusion model to generate more intricate texture details. Subsequently, we introduce a histogram-guided color consistency module that utilizes color histogram as guidance to learn precise color information for grayscale to sRGB color consistency mapping, with a color consistency loss designed to constrain the learned color information. Extensive experiments on publicly available real-world benchmarks show that the proposed ISPDiffuser outperforms state-of-the-art competitors both quantitatively and visually. Code will be released to facilitate future research.

ISPDiffuser: Learning RAW-to-sRGB Mappings with Texture-Aware Diffusion Models and Histogram-Guided Color Consistency

In clinical imaging, medical segmentation networks typically require continually adapting to new data from multiple sites over time, as aggregating all data for learning at once can be impractical due to storage limitations and privacy concerns. 
However, existing methods basically overlook domain-specific characteristics and fall short of adequately capturing domain-invariant knowledge during continual learning, leading to undesired catastrophic forgetting of previous sites and inferior generalization to new sites. 
To tackle this issue, this paper introduces FR2Seg, to sufficiently exploit both domain-specific and domain-invariant knowledge for efficient continual learning with the aid of low-frequency cues. 
For the former aspect, we propose a Fourier style reply module to synthesize pseudo images with old-site styles for data augmentation during new-site training, effectively preventing catastrophic forgetting without sacrificing data privacy. 
For the latter, we present a Fourier adaptive consistency regularization to identify and constrain the optimization of domain-invariant parameters with explicit awareness of knowledge transferability across sites, ensuring excellent generalizability to new sites. 
Experimental results on two public datasets confirm our method's superiority over existing state-of-the-art continual learning methods.

FR²Seg: Continual Segmentation Across Multiple Sites via Fourier Style Replay and Adaptive Consistency Regularization

Tsetlin Machines (TMs) have garnered increasing interest for their ability to learn concepts via propositional formulas and their proven efficiency across various application domains. Despite this, the convergence proof for the TMs, particularly for the AND operator (\emph{conjunction} of literals), in the generalized case (inputs greater than two bits) remains an open problem. This paper aims to fill this gap by presenting a comprehensive convergence analysis of Tsetlin automaton-based Machine Learning algorithms. We introduce a novel framework, referred to as Probabilistic Concept Learning (PCL), which simplifies the TM structure while incorporating dedicated feedback mechanisms and dedicated inclusion/exclusion probabilities for literals. Given $n$ features, PCL aims to learn a set of conjunction clauses $C_i$ each associated with a distinct inclusion probability $p_i$. Most importantly, we establish a theoretical proof confirming that, for any clause $C_k$, PCL converges to a conjunction of literals when $0.5<p_k<1$.
This result serves as a stepping stone for future research on the convergence properties of Tsetlin automaton-based learning algorithms. Our findings not only contribute to the theoretical understanding of Tsetlin automaton-based learning algorithms but also have implications for their practical application, potentially leading to more robust and interpretable machine learning models.

Generalized Convergence Analysis of Tsetlin Automaton Based Algorithms: A Probabilistic Approach to Concept Learning

Conditional independence (CI) testing is a fundamental task in modern statistics and machine learning. The conditional randomization test (CRT) was recently introduced to test whether two random variables, $X$ and $Y$, are conditionally independent given a potentially high-dimensional set of random variables, $Z$. The CRT operates exceptionally well under the assumption that the conditional distribution $X|Z$ is known. However, since this distribution is typically unknown in practice, accurately approximating it becomes crucial. In this paper, we propose using  conditional diffusion models (CDMs) to learn the distribution of $X|Z$. Theoretically and empirically, it is shown that CDMs  closely approximate the true conditional distribution. Furthermore, CDMs offer a more accurate approximation of  $X|Z$ compared to GANs, potentially leading to a CRT that performs better than those based on GANs. To accommodate complex dependency structures, we utilize a computationally efficient classifier-based conditional mutual information (CMI) estimator as our test statistic. The proposed testing procedure performs effectively without requiring assumptions about specific distribution forms or feature dependencies, 
and is capable of handling mixed-type conditioning sets that include both continuous and discrete variables. Theoretical analysis  shows that our proposed test achieves a valid control of the type I error. A series of experiments on synthetic data  demonstrates that our new test 
effectively controls both type-I and type-II errors,  even in  high dimensional scenarios.

Conditional Diffusion Models Based Conditional Independence Testing

Diffusion models have achieved remarkable success in the image and video generation tasks. Nevertheless, they often require a large amount of memory and time overhead during inference, due to the complex network architecture and considerable number of timesteps for iterative diffusion. Recently, the post-training quantization (PTQ) technique has proved a promising way to reduce the inference cost by quantizing the float-point operations to low-bit ones. However, most of them fail to tackle with the large variations in the distribution of activations across distinct channels and timesteps, as well as 
the inconsistent of input between quantization and inference on diffusion models, thus leaving much room for improvement. To address the above issues, we propose a novel method dubbed Timestep-Channel Adaptive Quantization for Diffusion Models (TCAQ-DM). Specifically, we develop a timestep-channel joint reparameterization (TCR) module to balance the activation range along both the timesteps and channels, facilitating the successive reconstruction procedure. Subsequently, we employ a dynamically adaptive quantization (DAQ) module that mitigate the quantization error by selecting an optimal quantizer for each post-Softmax layers according to their specific types of distributions. Moreover, we present a progressively aligned reconstruction (PAR) strategy to mitigate the bias caused by the input mismatch. Extensive experiments on various benchmarks and distinct diffusion models demonstrate that the proposed method substantially outperforms the state-of-the-art approaches in most cases, especially yielding comparable FID metrics to the full precision model  on CIFAR-10 in the W6A6 setting, while enabling generating available images in the W4A4 settings. We will release the source code upon acceptance.

TCAQ-DM: Timestep-Channel Adaptive Quantization for Diffusion Models

In recent years, large language models (LLMs) have developed rapidly and revolutionized natural language processing. However, high storage overhead and computing costs limit LLM deployment in resource-constrained environments. Quantization algorithms can effectively compress LLMs and accelerate inference, but they lead to loss in precision, especially in low-bit scenarios. In this paper, we find that the discarded weight values caused by quantization in fact contain treasures to improve LLMs' accuracy. To excavate those hidden treasures, we construct search spaces around these discarded weights and those weights within the search space can seamlessly be incorporated into the original quantization weight. To determine which weights should be merged, we design a plug-and-play weight compensation framework to capture global information and keep the weights with the highest potential benefits. Our framework can be combined with various LLM quantization algorithms to achieve higher precision without additional inference overhead. We validate the effectiveness of our approach on widely used benchmark datasets for LLMs. The code will be open-sourced after paper acceptance.

Treasures in Discarded Weights for LLM Quantization

Recent years have seen substantial progress in diffusion-based controllable video generation. 
However, achieving precise control in complex scenarios, including fine-grained object parts, sophisticated motion trajectories, and coherent background movement, remains a challenge. 
In this paper, we introduce *TrackGo*, a novel approach that leverages free-form masks and arrows for conditional video generation. 
This method offers users with a flexible and precise mechanism for manipulating video content.
We also propose the *TrackAdapter* for control implementation, an efficient and lightweight adapter designed to be seamlessly integrated into the temporal self-attention layers of a pretrained video generation model. This design leverages our observation that the attention map of these layers can accurately activate regions corresponding to motion in videos.
Our experimental results demonstrate that our new approach, enhanced by the TrackAdapter, achieves state-of-the-art performance on key metrics such as FVD, FID, and ObjMC scores.

TrackGo: A Flexible and Efficient Method for Controllable Video Generation

Talking head video generation involves animating a still face image using facial motion cues derived from a driving video to replicate target poses and expressions. Traditional methods often rely on the assumption that the relative positions of facial keypoints remain unchanged. However, this assumption fails when keypoints are occluded or when the head is in a profile pose, leading to inconsistencies in identity and blurring in certain facial regions.
In this paper, we introduce Occlusion-Insensitive Talking Head Video Generation, a novel approach that eliminates the reliance on spatial correlation of keypoints and instead leverages semantic correlation. Our method transforms facial features into a facelet semantic bank, where each facelet token represents a specific facial semantic. This bank is devoid of spatial information, allowing it to compensate for any invisible or occluded face regions during motion warping.
The facelet compensation module then populates the facelet tokens within the initially warped features by learning a correlation matrix between facial semantics and the facelet bank. This approach enables precise compensation for occlusions and pose changes, enhancing the fidelity of the generated videos.
Extensive experiments demonstrate that our method achieves state-of-the-art results, preserving source identity, maintaining fine-grained facial details, and capturing nuanced facial expressions with remarkable accuracy. Video demonstrations can be found in our supplementary materials.

Downloads

Next from AAAI 2025

K-hop Hypergraph Neural Network: A Comprehensive Aggregation Approach

Similar lecture

PTMQ: Post-training Multi-Bit Quantization of Neural Networks

.css-70qvj9{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;}Downloads

Next from AAAI 2025

K-hop Hypergraph Neural Network: A Comprehensive Aggregation Approach

Similar lecture

PTMQ: Post-training Multi-Bit Quantization of Neural Networks

Downloads