| Challenge: | Large pre-trained language models have demonstrated impressive capabilities, but there is still much to learn about how they operate. |
| Approach: | They investigate the ability of the autoregressive transformer to perform basic addition operations by using causal analysis to find that a few different attention heads in the middle layers control the addition carry . they found that due to the lack of global focus on the sequence within these attention heads, the model struggles to handle long-sequence addition tasks. |
| Outcome: | The model performs basic addition tasks, but it still faces challenges with length generalization. |
Similar Papers
Towards Incremental Transformers: An Empirical Analysis of Transformer Models for Incremental NLU (2021.emnlp-main)
Copied to clipboard
| Challenge: | Recent work attempts to apply incremental processing to NLUs but this is computationally expensive and does not scale efficiently for long sequences. |
| Approach: | They propose to apply Transformers incrementally via restart-incrementality by repeatedly feeding, to an unchanged model, increasingly longer input prefixes to produce partial outputs. |
| Outcome: | The proposed model has better incremental performance and faster inference speed compared to the standard Transformer and LT with restart-incrementality, at the cost of part of the non-incremental quality. |
Principled Understanding of Generalization for Generative Transformer Models in Arithmetic Reasoning Tasks (2025.acl-long)
Copied to clipboard
| Challenge: | Existing models excel in arithmetic reasoning but their generalization capabilities are incompletely understood. |
| Approach: | They propose a theoretical framework for understanding the generalization behaviors of transformers in arithmetic tasks, focusing on length generalization. |
| Outcome: | The proposed framework can predict generalization behaviors in transformers with a high translation invariance and base mismatch in modular operations. |
Efficient Long-Range Transformers: You Need to Attend More, but Not Necessarily at Every Layer (2023.findings-emnlp)
Copied to clipboard
| Challenge: | Pretrained transformer models have demonstrated remarkable performance across various natural language processing tasks. |
| Approach: | They propose a transformer variant with mixed attention spans that leverages the attention mechanism to capture long- and short-range dependencies in the sequence. |
| Outcome: | The proposed model can achieve competitive performance to models with full attention while reducing computational cost (75%) |
The NLP Task Effectiveness of Long-Range Transformers (2023.eacl-main)
Copied to clipboard
| Challenge: | Existing benchmarks on long-range attention models have not been sufficient to develop efficient Transformers and their practical application on complex NLP tasks. |
| Approach: | They propose to benchmark 7 Transformer variants on 5 difficult NLP tasks and 7 datasets to examine their capacity for long-range attention. |
| Outcome: | The proposed models have advantages on content selection and query-guided decoding, but they come with previously unrecognized drawbacks such as insufficient attention to distant tokens and accumulated approximation error. |
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. |
How Numerical Precision Affects Arithmetical Reasoning Capabilities of LLMs (2025.findings-acl)
Copied to clipboard
Guhao Feng, Kai Yang, Yuntian Gu, Xinyue Ai, Shengjie Luo, Jiacheng Sun, Di He, Zhenguo Li, Liwei Wang
| Challenge: | Despite the success of transformer-based large language models, understanding and enhancing their mathematical capabilities remains a significant challenge. |
| Approach: | They propose to use numerical precision as a key factor that influences LLMs' effectiveness in arithmetical tasks to determine their effectiveness. |
| Outcome: | The proposed models perform better in arithmetic tasks than transformer-based models with standard numerical precision. |
Fine-Tuning Pre-trained Transformers into Decaying Fast Weights (2022.emnlp-main)
Copied to clipboard
| Challenge: | Autoregressive Transformers incur O(T) complexity during per-token generation due to the self-attention mechanism. |
| Approach: | They propose a kernel-based method to approximate causal self-attention by replacing it with recurrent formulations with various update rules and feature maps to achieve O(1) time and memory complexity. |
| Outcome: | The proposed method outperforms prior methods and retains 99% of attention’s performance on WikiText-103 against more complex attention substitutes. |
Analyzing the Inner Workings of Transformers in Compositional Generalization (2025.naacl-long)
Copied to clipboard
| Challenge: | Existing studies on compositional generalization abilities of neural models have focused on benchmarks, but the results do not reflect the underlying competence of the model. |
| Approach: | They propose to find an existing subnetwork that contributes to the generalization performance and perform causal analyses on how the model utilizes syntactic features. |
| Outcome: | The proposed model relies on syntactic features but the subnetwork with better generalization performance relies mainly on a non-compositional algorithm . |
Attention Alignment and Flexible Positional Embeddings Improve Transformer Length Extrapolation (2024.findings-naacl)
Copied to clipboard
| Challenge: | Existing methods for length extrapolation are tailored for natural language modeling, a task known to have strong recency bias. |
| Approach: | They propose two attention alignment strategies to improve T5's long-context utilization capability without fine-tuning. |
| Outcome: | The proposed methods improve the long-context utilization capability of T5 on language modeling, retrieval, multi-document question answering, and code completion tasks without any fine-tuning. |
Shortformer: Better Language Modeling using Shorter Inputs (2021.acl-long)
Copied to clipboard
| Challenge: | Existing methods require computationally expensive relative position embeddings. |
| Approach: | They propose two methods that decrease input length to improve perplexity and perplexability. |
| Outcome: | The proposed methods speed up training by a factor of 1.65 and reduce memory usage. |