7 citations · 17 across the 21 of their papers we have counts for
22 papers
Unlocking Lossless Speedups in LLMs via Discrete Diffusion
Subham Sekhar Sahoo, Lingjie Chen, Khiem Pham +14
Large Language Models (LLMs) owe much of their success to next-token prediction (NTP), but their autoregressive (AR) structure requires slow, sequential token generation. To overco…
Compute Where it Counts: Self Optimizing Language Models
Yash Akhauri, Mohamed S. Abdelfattah
Efficient LLM inference research has largely focused on reducing the cost of each decoding step (e.g., using quantization, pruning, or sparse attention), typically applying a unifo…
Regression Language Models for Code
Yash Akhauri, Xingyou Song, Arissa Wongpanich +2
We study code-to-metric regression: predicting numeric outcomes of code executions, a challenging task due to the open-ended nature of programming languages. While prior methods ha…
Performance Prediction for Large Systems via Text-to-Text Regression
Yash Akhauri, Bryan Lewandowski, Cheng-Hsi Lin +7
In many industries, predicting metric outcomes of large systems is a fundamental problem, driven largely by traditional tabular regression. However, such methods struggle on comple…
Esoteric Language Models: A Family of Any-Order Diffusion LLMs
Subham Sekhar Sahoo, Zhihan Yang, Yash Akhauri +7
Diffusion-based language models offer a compelling alternative to autoregressive (AR) models by enabling parallel and controllable generation. Within this family, Masked Diffusion…
FlashDLM: Accelerating Diffusion Language Model Inference via Efficient KV Caching and Guided Diffusion
Zhanqiu Hu, Jian Meng, Yash Akhauri +4
Diffusion language models offer parallel token generation and inherent bidirectionality, promising more efficient and powerful sequence modeling compared to autoregressive approach…
SplitReason: Learning To Offload Reasoning
Yash Akhauri, Anthony Fei, Chi-Chih Chang +3
Reasoning in large language models (LLMs) tends to produce substantially longer token generation sequences than simpler language modeling tasks. This extended generation length ref…
xKV: Cross-Layer KV-Cache Compression via Aligned Singular Vector Extraction
Chi-Chih Chang, Wei-Cheng Lin, Chien-Yu Lin +8
Long-context Large Language Models (LLMs) enable powerful applications but incur high memory costs due to the key-value states (KV-Cache). Recent studies attempt to share KV-Cache…
TokenButler: Token Importance is Predictable
Yash Akhauri, Ahmed F AbouElhamayed, Yifei Gao +4
Large Language Models (LLMs) rely on the Key-Value (KV) Cache to store token history, enabling efficient decoding of tokens. As the KV-Cache grows, it becomes a major memory and co…
SparAMX: Accelerating Compressed LLMs Token Generation on AMX-powered CPUs
Ahmed F. AbouElhamayed, Jordan Dotzel, Yash Akhauri +6
Large language models have high compute, latency, and memory requirements. While specialized accelerators such as GPUs and TPUs typically run these workloads, CPUs are more widely…
Attamba: Attending To Multi-Token States
Yash Akhauri, Safeen Huda, Mohamed S. Abdelfattah
When predicting the next token in a sequence, vanilla transformers compute attention over all previous tokens, resulting in quadratic scaling of compute with sequence length. State…
ShadowLLM: Predictor-based Contextual Sparsity for Large Language Models
Yash Akhauri, Ahmed F AbouElhamayed, Jordan Dotzel +4
The high power consumption and latency-sensitive deployments of large language models (LLMs) have motivated efficiency techniques like quantization and sparsity. Contextual sparsit…
Radial Networks: Dynamic Layer Routing for High-Performance Large Language Models
Jordan Dotzel, Yash Akhauri, Ahmed S. AbouElhamayed +3
Large language models (LLMs) often struggle with strict memory, latency, and power demands. To meet these demands, various forms of dynamic sparsity have been proposed that reduce…
Encodings for Prediction-based Neural Architecture Search
Yash Akhauri, Mohamed S. Abdelfattah
Predictor-based methods have substantially enhanced Neural Architecture Search (NAS) optimization. The efficacy of these predictors is largely influenced by the method of encoding…
On Latency Predictors for Neural Architecture Search
Yash Akhauri, Mohamed S. Abdelfattah
Efficient deployment of neural networks (NN) requires the co-optimization of accuracy and latency. For example, hardware-aware neural architecture search has been used to automatic…
Multi-Predict: Few Shot Predictors For Efficient Neural Architecture Search
Yash Akhauri, Mohamed S. Abdelfattah
Many hardware-aware neural architecture search (NAS) methods have been developed to optimize the topology of neural networks (NN) with the joint objectives of higher accuracy and l…
EZNAS: Evolving Zero Cost Proxies For Neural Architecture Scoring
Yash Akhauri, J. Pablo Munoz, Nilesh Jain +1
Neural Architecture Search (NAS) has significantly improved productivity in the design and deployment of neural networks (NN). As NAS typically evaluates multiple models by trainin…
Enabling NAS with Automated Super-Network Generation
J. Pablo Muñoz, Nikolay Lyalyushkin, Yash Akhauri +3
Recent Neural Architecture Search (NAS) solutions have produced impressive results training super-networks and then deriving subnetworks, a.k.a. child models that outperform expert…
RHNAS: Realizable Hardware and Neural Architecture Search
Yash Akhauri, Adithya Niranjan, J. Pablo Muñoz +6
The rapidly evolving field of Artificial Intelligence necessitates automated approaches to co-design neural network architecture and neural accelerators to maximize system efficien…
Exposing Hardware Building Blocks to Machine Learning Frameworks
Yash Akhauri
There are a plethora of applications that demand high throughput and low latency algorithms leveraging machine learning methods. This need for real time processing can be seen in i…
LogicNets: Co-Designed Neural Networks and Circuits for Extreme-Throughput Applications
Yaman Umuroglu, Yash Akhauri, Nicholas J. Fraser +1
Deployment of deep neural networks for applications that require very high throughput or extremely low latency is a severe computational challenge, further exacerbated by inefficie…
HadaNets: Flexible Quantization Strategies for Neural Networks
Yash Akhauri
On-board processing elements on UAVs are currently inadequate for training and inference of Deep Neural Networks. This is largely due to the energy consumption of memory accesses i…