Papers with resource-constrained devices
Extreme Model Compression for On-device Natural Language Understanding (2020.coling-industry)
Copied to clipboard
| Challenge: | Xu and Sarikaya et al., 2014) perform word-embedding compression with NLU task learning . their approach achieves a compression rate of 97.4% with less than 3.7% degradation in predictive performance. |
| Approach: | They propose a task-aware, end-to-end compression approach that performs word-embedding compression with NLU task learning. |
| Outcome: | The proposed approach outperforms baselines and achieves 97.4% compression rate with less than 3.7% degradation in predictive performance. |
Building an Efficient Multilingual Non-Profit IR System for the Islamic Domain Leveraging Multiprocessing Design in Rust (2024.emnlp-industry)
Copied to clipboard
| Challenge: | Existing models that are pre-trained on a general domain can deteriorate performance due to domain shift when applied to new domains. |
| Approach: | They propose to train a multilingual non-profit IR system for the Islamic domain using Rust Language capabilities. |
| Outcome: | The proposed model outperforms models pre-trained on general domains and on resource-constrained devices. |
COST-EFF: Collaborative Optimization of Spatial and Temporal Efficiency with Slenderized Multi-exit Language Models (2022.emnlp-main)
Copied to clipboard
| Challenge: | Existing statically compressed pre-trained language models lack spatial and temporal efficiency due to their large size and wide width. |
| Approach: | They propose a spatially and temporally efficient model which retains the major capacity of PLMs. |
| Outcome: | The proposed model retains the major capacity of pre-trained language models at high compression and acceleration rate with 1/8 parameters and 1/19 FLOPs of BERT. |
AutoTinyBERT: Automatic Hyper-parameter Optimization for Efficient Pre-trained Language Models (2021.acl-long)
Copied to clipboard
| Challenge: | Pre-trained language models (PLMs) have achieved great success in natural language processing. |
| Approach: | They propose a method that automatically searches architecture hyper-parameters in BERT . they use one-shot learning and the search space to provide an adaptive development way . |
| Outcome: | The proposed method outperforms both the baseline and distillation-based methods on GLUE and SQUAD benchmarks. |
MLWQ: Efficient Small Language Model Deployment via Multi-Level Weight Quantization (2025.emnlp-main)
Copied to clipboard
| Challenge: | Existing methods for efficient deployment of small language models face inefficient bit-width allocation and insufficient fine-grained quantization adjustments. |
| Approach: | They propose a weight quantization technique that facilitates efficient deployment of SLMs . they propose to combine inter-layer loss and intra-layer salience to achieve better allocation . |
| Outcome: | Experimental results show that multi-level weight quantization achieves competitive performance compared to state-of-the-art methods. |
Revisiting Pruning vs Quantization for Small Language Models (2025.findings-emnlp)
Copied to clipboard
| Challenge: | Compressing Small Language Models (SLMs) is particularly suited for resource-constrained devices, but their compression dynamics remain underexplored compared to Large Language Model (LLMs). |
| Approach: | They evaluated post-training pruning and quantization methods across six SLMs from 0.5 to 3.8B, seven languages, and seven downstream tasks. |
| Outcome: | The proposed methods outperform pruning and quantization on six SLMs from 0.5 to 3.8B, seven languages, and seven downstream tasks. |
Demystifying Small Language Models for Edge Deployment (2025.acl-long)
Copied to clipboard
Zhenyan Lu, Xiang Li, Dongqi Cai, Rongjie Yi, Fangming Liu, Wei Liu, Jian Luan, Xiwen Zhang, Nicholas D. Lane, Mengwei Xu
| Challenge: | Small language models (SLMs) are a promising solution for resource-constrained devices such as smartphones and the Web of Things. |
| Approach: | They propose to use SLMs to build and optimize a set of small language models that are publicly accessible. |
| Outcome: | The proposed models outperform 7B models in general tasks, while their in-context learning capabilities remain limited and their efficiency has significant optimization potential. |
Differentially Private Knowledge Distillation via Synthetic Text Generation (2024.findings-acl)
Copied to clipboard
| Challenge: | Large Language models (LLMs) are achieving state-of-the-art performance in many downstream tasks, but data privacy is a major challenge for practitioners. |
| Approach: | They propose a differentially private knowledge distillation algorithm that exploits the knowledge of a teacher LLM and a student's output distribution. |
| Outcome: | The proposed algorithm significantly improves the utility over baselines on the Big Patent dataset, with strong privacy parameters, =2. |
Speculative Streaming: Efficient and Scalable Speculative Decoding with Multi-Stream Attention (2025.emnlp-main)
Copied to clipboard
Nikhil Bhendawade, Irina Belousova, Qichen Fu, Henry Mason, Antonie Lin, Mohammad Rastegari, Mahyar Najibi
| Challenge: | Speculative decoding is a prominent technique for accelerating LLM inference by leveraging an auxiliary draft model, but its effectiveness is limited by the autoregressive nature of draft generation. |
| Approach: | They propose a method that integrates speculative draft generation directly within the target model using multi-stream attention. |
| Outcome: | The proposed method improves acceptance but also latency and speculation latency, limiting overall speedup. |
FAQ: Mitigating Quantization Error via Regenerating Calibration Data with Family-Aware Quantization (2026.findings-acl)
Copied to clipboard
| Challenge: | representativeness and universality of calibration data remain a bottleneck in quantization accuracy. |
| Approach: | They propose a framework that leverages prior knowledge from LLMs to generate calibration samples . their framework reduces accuracy loss by up to 28.5% compared to baseline . |
| Outcome: | Experiments show that family-aware quantization reduces accuracy loss by up to 28.5% compared to baseline data. |