Attention Entropy is a Key Factor: An Analysis of Parallel Context Encoding with Full-attention-based Pre-trained Language Models (2025.acl-long)
Copied to clipboard
Zhisong Zhang, Yan Wang, Xinting Huang, Tianqing Fang, Hongming Zhang, Chenlong Deng, Shuaiyi Li, Dong Yu
| Challenge: | Large language models have demonstrated remarkable performance across a wide range of language tasks due to their remarkable ability in context modeling. |
| Approach: | They propose to use parallel context encoding to reduce attention entropy by incorporating attention sinks and selective mechanisms to reduce irregular attention . they also propose to incorporate attention sink mechanisms into the parallel encoded context to reduce the irregular attention. |
| Outcome: | The proposed methods lower irregular attention entropy and narrow performance gaps. |
Similar Papers
Entropy-Based Decoding for Retrieval-Augmented Large Language Models (2025.naacl-long)
Copied to clipboard
| Challenge: | Despite their success, retrieval-augmented LLMs still face the distractibility issue, where the generated responses are negatively influenced by noise from both external and intrinsic knowledge sources. |
| Approach: | They propose a entropy-based document-parallel ensemble decoding method that prioritizes low-entropies from retrieved documents and incorporates a contrastive decoding mechanism that contrasts the obtained low- and high-entropic ensemble distributions with the high-end internal knowledge across layers. |
| Outcome: | The proposed method improves on open-domain question answering datasets and shows that it is highly efficient. |
Dynamic Attention-Guided Context Decoding for Mitigating Context Faithfulness Hallucinations in Large Language Models (2025.findings-acl)
Copied to clipboard
| Challenge: | Existing methods, such as a n-terminal coding, do not provide accurate data for large language models. |
| Approach: | They propose a lightweight framework that leverages attention distributions and uncertainty signals in a single-pass decoding. |
| Outcome: | Experiments on open-book QA datasets show that DAGCD improves faithfulness and robustness while preserving computational efficiency. |
Variance Sensitivity Induces Attention Entropy Collapse and Instability in Transformers (2025.emnlp-main)
Copied to clipboard
| Challenge: | Attention-based language models rely on the softmax function to convert attention logits into probability distributions, but this process can result in attention entropy collapse. |
| Approach: | They propose to use the softmax function to re-weight attention logits to create probability distributions, but this reweighting can lead to attention entropy collapse . they find that entropic-stable attention methods can prevent entrapment and enable more stable training by controlling or insensitive to variance of attention logit variance. |
| Outcome: | The proposed methods prevent attention entropy collapse and enable more stable training. |
Exploring Attention Attractors in Large Language Models (2026.acl-long)
Copied to clipboard
| Challenge: | Existing studies have suggested that attention attractors function as "summary tokens" while others speculate that tokens with weaker semantics attract high attention, they act as attention sinks that offload excessive attention. |
| Approach: | They examine attention attractors, tokens that draw significantly high attention, in large language models. |
| Outcome: | The proposed models are able to capture long-range dependencies within a given context. |
Rethinking the Role of Scale for In-Context Learning: An Interpretability-based Case Study at 66 Billion Scale (2023.acl-long)
Copied to clipboard
| Challenge: | 70% of attention heads and 20% of the feed forward networks can be removed with minimal decline in task performance. |
| Approach: | They propose to investigate whether in-context learning is not uniform across all components of a large language model. |
| Outcome: | The proposed model can remove 70% of attention heads and 20% of feed forward networks with minimal decline in task performance. |
Data-Informed Global Sparseness in Attention Mechanisms for Deep Neural Networks (2024.lrec-main)
Copied to clipboard
| Challenge: | Attention pruning techniques have been developed to identify and exploit sparseness . previous work has taken pioneering steps to discover and explain the sparsity in attention patterns . |
| Approach: | They propose a framework that observes attention patterns in a fixed dataset and generates a global sparseness mask. |
| Outcome: | The proposed approach saves 90% of computations and maintains quality of results. |
Recurrent Attention Networks for Long-text Modeling (2023.findings-acl)
Copied to clipboard
| Challenge: | Existing approaches to encoding long documents using self-attention have been limited by quadratic computational complexities and limited application in long text processing. |
| Approach: | They propose a long-document encoding model that allows the recurrent operation of self-attention. |
| Outcome: | The proposed model extracts global semantics in token-level and document-level representations, making it inherently compatible with both sequential and sequential tasks. |
Attention Can Reflect Syntactic Structure (If You Let It) (2021.eacl-main)
Copied to clipboard
| Challenge: | a recent study has attempted to decode linguistic structure from the Transformer . but, much of the work focused on English, a language with rigid word order and a lack of inflectional morphology. |
| Approach: | They propose to fine-tune a feature encoder for BERT to learn linguistic structure from its multi-head attention mechanism. |
| Outcome: | The proposed model can decode full trees above baseline accuracy from single attention heads across languages. |
On-the-Fly Attention Modulation for Neural Generation (2021.findings-acl)
Copied to clipboard
Yue Dong, Chandra Bhagavatula, Ximing Lu, Jena D. Hwang, Antoine Bosselut, Jackie Chi Kit Cheung, Yejin Choi
| Challenge: | Degeneration of neural text is associated with insufficient learning of task-specific characteristics by the attention mechanism. |
| Approach: | They propose to use attention modulation to inject priors into inference to improve fluency, creativity, and commonsense reasoning in neural text generation models. |
| Outcome: | The proposed method improves fluency, creativity, and commonsense reasoning, and significantly reduces sentence-level repetition. |
A Silver Bullet or a Compromise for Full Attention? A Comprehensive Study of Gist Token-based Context Compression (2025.acl-long)
Copied to clipboard
| Challenge: | gist-based context compression methods can achieve only slight performance loss on tasks like retrieval-augmented generation and long-document QA, but it faces challenges in tasks like synthetic recall. |
| Approach: | They propose two strategies to improve gist-based context compression in large language models. |
| Outcome: | The proposed methods can achieve only slight performance loss on retrieval-augmented generation and long-document QA tasks, but they face challenges in tasks like synthetic recall. |