On Sparsifying Encoder Outputs in Sequence-to-Sequence Models (2021.findings-acl)
Copied to clipboard
| Challenge: | Using sequence-to-sequence models, encoder outputs are usually transferred to the decoder for generation, but in this study, encoded outputs can be compressed to shorten the sequence for decoding. |
| Approach: | They propose to use a stochastic gate-based algorithm to mask encoder outputs to shorten the sequence delivered for decoding. |
| Outcome: | The proposed model can be used to shorten encoder outputs to short a sequence . the proposed model yields a speedup of up to 1.65 on document summarization and 1.20 on character-based machine translation tasks. |
Similar Papers
Sparsity and Sentence Structure in Encoder-Decoder Attention of Summarization Systems (2021.emnlp-main)
Copied to clipboard
| Challenge: | Training and inference using large transformer models can be computationally expensive because the self-attention's time and memory grow quadratically with sequence length. |
| Approach: | They propose a modified transformer architecture that constrains the encoder-decoder attention mechanism to a subset of input sentences while maintaining system performance. |
| Outcome: | The proposed architecture can be trained and inferenced using large transformer models with expensive training and induction costs. |
FiRST: Finetuning Router-Selective Transformers for Input-Adaptive Latency Reduction (2025.findings-emnlp)
Copied to clipboard
| Challenge: | Existing approaches to improve latency via skipping layers have limitations . fiRST is a model-agnostic framework that reduces inference latency while maintaining quality . |
| Approach: | They propose a model-agnostic framework that skips transformer layers during decoding . it is fully compatible with KV caching, enabling faster decoding while maintaining quality . |
| Outcome: | a new framework reduces inference latency by using layer-specific routers to skip transformer layers during decoding. |
Modular Transformers: Compressing Transformers into Modularized Layers for Flexible Efficient Inference (2023.findings-acl)
Copied to clipboard
| Challenge: | Pre-trained sequence-to-sequence models have advanced the state of the art on text generation tasks. |
| Approach: | They introduce a modular encoder-decoder framework for flexible sequence-to-sequence model compression. |
| Outcome: | The proposed framework can achieve flexible compression ratios from 1.1x to 6x with little to moderate relative performance drop. |
SparseFlow: Accelerating Transformers by Sparsifying Information Flows (2024.acl-long)
Copied to clipboard
| Challenge: | SparseFlow is an efficient method to sparsify the dense information flows within transformers. |
| Approach: | They propose a method to sparsify the dense pathways of token representations across all transformer blocks by parameterizing them to be sparse. |
| Outcome: | The proposed method reduces computational costs by half on average without compromising task accuracy. |
Sparsifying Transformer Models with Trainable Representation Pooling (2022.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to sparsify attention in the Transformer model are based on quadratic memory complexity and a lack of information for each word. |
| Approach: | They propose a method to sparsify attention in a Transformer model by learning to select the most-informative token representations during the training process. |
| Outcome: | The proposed model performs better than the current SOTA model while being 1.8 faster during training, 4.5 faster inference and 13 more efficient in the decoder. |
Efficient Contextualized Representation: Language Model Pruning for Sequence Labeling (D18-1)
Copied to clipboard
| Challenge: | Existing efforts to train pre-trained language models have brought significant improvements to various NLP applications. |
| Approach: | They propose to compress bulky LMs while preserving useful information for a specific task. |
| Outcome: | The proposed method can detach any layer without affecting others, and stretch shallow and wide LMs to be deep and narrow. |
SEQˆ3: Differentiable Sequence-to-Sequence-to-Sequence Autoencoder for Unsupervised Abstractive Sentence Compression (N19-1)
Copied to clipboard
| Challenge: | Neural sequence-to-sequence models are currently the dominant approach in natural language processing tasks, but require massive parallel corpora. |
| Approach: | They propose a sequence-to-sequence-tosequnce autoencoder with words as latent variables . they apply the model to unsupervised abstractive sentence compression . |
| Outcome: | The proposed model achieves promising results in unsupervised sentence compression on benchmark datasets. |
Bringing Emerging Architectures to Sequence Labeling in NLP (2026.eacl-long)
Copied to clipboard
| Challenge: | Pretrained Transformer encoders are the dominant approach to sequence labeling . however, few have been applied to sequence labels on flat or simplified tasks . |
| Approach: | They propose to use pretrained Transformer encoders to model relations across words . they find that the architectures adapt well across tagging tasks that vary in complexity . |
| Outcome: | The proposed architectures perform well across tagging tasks across languages and datasets. |
Sparse Sequence-to-Sequence Models (P19-1)
Copied to clipboard
| Challenge: | Sequence-to-sequence models are dense and assigning nonzero probability to implausible outputs. |
| Approach: | They propose a new family of -entmax transformations that includes softmax and sparsemax as particular cases and is sparser for any > 1 . they provide fast algorithms to evaluate these transformations and their gradients, which scale well for large vocabulary sizes. |
| Outcome: | The proposed models are able to produce sparse alignments and assign nonzero probability to short list of plausible outputs, sometimes rendering beam search exact. |
On the Distribution, Sparsity, and Inference-time Quantization of Attention Values in Transformers (2021.findings-acl)
Copied to clipboard
Tianchu Ji, Shraddhan Jain, Michael Ferdman, Peter Milder, H. Andrew Schwartz, Niranjan Balasubramanian
| Challenge: | Recent work shows that attention can be pruned to zeros with minimal loss in accuracy. |
| Approach: | They propose a pruning technique which quantizes attention to a 3-bit format without retraining . they find that 80% of attention values can be pruned to zeros with minimal loss in accuracy . |
| Outcome: | The proposed approach produces only a few unique attention values with minimal loss in accuracy. |