Papers by Sainbayar Sukhbaatar
Training Hybrid Language Models by Marginalizing over Segmentations (P19-1)
Copied to clipboard
| Challenge: | Statistical language modeling is the problem of estimating a probability distribution over text data. |
| Approach: | They propose to marginalize over the segmentations efficiently to compute the true probability of a sequence. |
| Outcome: | The proposed model marginalizes over the segmentations to compute the true probability of a sequence on three datasets comprising seven languages. |
Meta-Rewarding Language Models: Self-Improving Alignment with LLM-as-a-Meta-Judge (2025.emnlp-main)
Copied to clipboard
Tianhao Wu, Weizhe Yuan, Olga Golovneva, Jing Xu, Yuandong Tian, Jiantao Jiao, Jason E Weston, Sainbayar Sukhbaatar
| Challenge: | Existing methods for improving large language models have focused on improving model responses rather than judgment capabilities, resulting in rapid saturation during iterative training. |
| Approach: | They propose an iterative Meta-Rewarding step where the model judges its own judgements and uses that feedback to refine its judgment skills. |
| Outcome: | The proposed model improves Llama-3-8B-Instruct from 22.9% to 39.4% on AlpacaEval 2 and 20.6% to 29.1% on Arena-Hard. |
Adaptive Attention Span in Transformers (P19-1)
Copied to clipboard
| Challenge: | We extend the maximum context size of a neural network called Transformer to 8k characters. |
| Approach: | They propose a self-attention mechanism that can learn its optimal attention span . this allows for models with longer context and the capability to catch longer dependencies. |
| Outcome: | The proposed model achieves state-of-the-art performance on text8 and enwiki8 using 8k characters with no loss of performance, and maintains control over memory footprint and computational time. |
The CRINGE Loss: Learning what language not to model (2023.acl-long)
Copied to clipboard
| Challenge: | Standard language model training uses gold human documents or human-human interaction data and treats all training data as positive examples. |
| Approach: | They propose a procedure to train with negative examples using the "CRINGE" loss technique and use it to train models with such data. |
| Outcome: | The proposed procedure outperforms multiple strong baselines and is simple to train and implement. |
Following Length Constraints in Instructions (2025.emnlp-main)
Copied to clipboard
| Challenge: | Existing instruction following models fail to follow length constraints in their evaluations. |
| Approach: | They propose to train models that can be controlled at inference time with instructions containing desired length constraints. |
| Outcome: | The proposed models outperform standard instruction following models in length instructed evaluations. |
Director: Generator-Classifiers For Supervised Language Modeling (2022.aacl-main)
Copied to clipboard
| Challenge: | Current language models achieve low perplexity but their resulting generations still suffer from toxic responses, repetitiveness, and contradictions. |
| Approach: | They propose a new language model architecture that uses a language modeling and a classification head for each output token. |
| Outcome: | The proposed model outperforms existing model guiding approaches in terms of accuracy and efficiency. |