Upsample or Upweight? Balanced Training on Heavily Imbalanced Datasets (2025.naacl-long)
Copied to clipboard
| Challenge: | a lack of data across domains creates significant imbalances in training data sizes . a recent study shows that temperature sampling and scaling are equivalent but differ under stochastic gradient descent due to differences in gradient variance. |
| Approach: | They propose a method that upsamples low-resource languages and upweights their loss functions to address this disparity. |
| Outcome: | The proposed method competes effectively with existing data re-weighting techniques while offering computational efficiency. |
Similar Papers
Rethinking Data Mixture for Large Language Models: A Comprehensive Survey and New Perspectives (2026.findings-eacl)
Copied to clipboard
| Challenge: | Existing methods for data mixture improve the generalization capability of large language models (LLMs) on downstream tasks. |
| Approach: | They propose a fine-grained categorization of existing methods and propose three subtypes of offline and online methods. |
| Outcome: | The proposed methods extend beyond offline and online classifications and highlight key challenges in the field of data mixture. |
Learning Is Not A Race: Improving Retrieval in Language Models via Equal Learning (2025.findings-emnlp)
Copied to clipboard
| Challenge: | Overparametrized models trained on cross-entropy loss can overfit on noise . Fitting some tokens early reduces gradient signals in later iterations . |
| Approach: | They propose to overfit models trained on cross-entropy loss on noise . fitting some tokens early reduces gradient signals in later iterations . |
| Outcome: | The proposed approaches can be applied to large language models with longer contexts or larger embedding sizes. |
Multilingual Encoder Knows more than You Realize: Shared Weights Pretraining for Extremely Low-Resource Languages (2025.acl-long)
Copied to clipboard
| Challenge: | XLM-R and mBART have advanced multilingualism in NLP, but low-resource languages such as Tibetan, Uyghur, Kazakh, and Mongolian are underserved. |
| Approach: | They propose a framework for adapting multilingual encoders to text generation in extremely low-resource languages by reusing the weights between the encoder and the decoder. |
| Outcome: | The proposed framework performs better on various downstream tasks even when compared with much larger models. |
Revisiting Scaling Laws for Language Models: The Role of Data Quality and Training Strategies (2025.acl-long)
Copied to clipboard
| Challenge: | Existing scaling laws suggest augmenting model size and training data results in enhanced performance, but recent studies reveal deviations, particularly in large language models, where performance improvements decelerate—a phenomenon known as sub-scaling. |
| Approach: | They propose a sub-optimal scaling law that better predicts performance in sub-scaling regimes by examining data quality and training strategies. |
| Outcome: | The proposed scaling law better predicts performance in sub-scaling regimes, highlighting the importance of data quality and diversity. |
Tricks for Training Sparse Translation Models (2022.naacl-main)
Copied to clipboard
| Challenge: | Multitask learning with an unbalanced data distribution skews model learning towards high resource tasks. |
| Approach: | They propose to use a temperature heating mechanism and dense pre-training to mitigate this by training models with a fixed model capacity. |
| Outcome: | The proposed techniques improve performance on two multilingual translation benchmarks compared to BASELayers and Dense scaling baselines and in combination, more than 2x model convergence speed. |
DRPruning: Efficient Large Language Model Pruning through Distributionally Robust Optimization (2025.acl-long)
Copied to clipboard
| Challenge: | Structured pruning reduces model size but often causes uneven degradation across domains, leading to biased performance. |
| Approach: | They propose a method that dynamically adjusts the data distribution during training to restore balanced performance across heterogeneous and multi-tasking data. |
| Outcome: | Experiments in monolingual and multilingual settings show that the proposed method surpasses similarly sized models in pruning and continued pretraining over perplexity, downstream tasks, and instruction tuning. |
Balancing Training for Multilingual Neural Machine Translation (2020.acl-main)
Copied to clipboard
| Challenge: | Existing methods to train multilingual machine translation models are imbalanced and heterogeneous data is wildly varying. |
| Approach: | They propose a method that automatically learns how to weight training data through a data scorer that is optimized to maximize performance on all test languages. |
| Outcome: | The proposed method outperforms baselines on two sets of languages under one-to-many and many-to-1 MT settings and offers flexible control over which languages are optimized. |
High-quality Data-to-Text Generation for Severely Under-Resourced Languages with Out-of-the-box Large Language Models (2024.findings-eacl)
Copied to clipboard
| Challenge: | Pretrained large language models (LLMs) can bridge the performance gap for under-resourced languages by substantial margins, as measured by both automatic and human evaluations. |
| Approach: | They propose to use pretrained large language models to bridge this gap by automating and evaluating data-to-text generation in under-resourced languages. |
| Outcome: | The proposed model can set the state of the art for under-resourced languages by substantial margins, as measured by both automatic and human evaluations. |
Perturbation Augmentation for Fairer NLP (2022.emnlp-main)
Copied to clipboard
| Challenge: | Unwanted and often harmful social biases are becoming more salient in NLP research. |
| Approach: | They propose to train a neural perturbation model that rewrites demographic references in text to make them more fair. |
| Outcome: | The proposed model outperforms heuristic alternatives on a large dataset of human annotated text perturbations. |
Building Better: Avoiding Pitfalls in Developing Language Resources when Data is Scarce (2025.acl-long)
Copied to clipboard
| Challenge: | Language is a powerful means of communication and should be regarded as more than just a collection of tokens. |
| Approach: | They collect feedback from individuals directly involved in and impacted by NLP artefacts for medium- and low-resource languages and highlight key issues related to data quality, cultural appropriateness and ethics of common annotation practices. |
| Outcome: | The findings highlight key issues related to data quality, cultural appropriateness, and ethics of common annotation practices. |