Papers by William Chan
Non-Autoregressive Machine Translation with Latent Alignments (2020.emnlp-main)
Copied to clipboard
| Challenge: | Existing non-autoregressive machine translation methods are lacking in the field of latent alignments. |
| Approach: | They propose two strong methods for non-autoregressive machine translation that model latent alignments with dynamic programming. |
| Outcome: | The proposed models achieve state-of-the-art on the WMT’14 EnDe task, compared with the autoregressive Transformer baseline. |
An Empirical Study of Generation Order for Machine Translation (2020.emnlp-main)
Copied to clipboard
| Challenge: | a recent study of generation order for machine translation shows it does not affect output quality . Neural sequence models have been successfully applied to a broad range of tasks in recent years . |
| Approach: | They propose a soft order-reward framework that enables models to follow arbitrary oracle generation policies. |
| Outcome: | The proposed framework explores a wide variety of generation orders including uninformed orders, location-based orders, frequency-based or model-based orderings, and model-driven orders. |
Big Bidirectional Insertion Representations for Documents (D19-56)
Copied to clipboard
| Challenge: | Recent studies suggest we are nearing human-level parity for sentence-level translation in certain domains. |
| Approach: | They propose an insertion-based model for document-level translation tasks that embeds sentence alignment between the source and target document. |
| Outcome: | The proposed model improves on the WMT’19 English->German translation task by +4.3 BLEU compared with the Insertion Transformer baseline. |
Illustrative Language Understanding: Large-Scale Visual Grounding with Image Search (P18-1)
Copied to clipboard
| Challenge: | a large-scale lookup operation to ground language via ‘snapshots’ of our physical world accessed through image search is currently used to learn word representations. |
| Approach: | They propose a large-scale lookup operation to ground language via ‘snapshots’ of our physical world accessed through image search. |
| Outcome: | The proposed model is based on a large-scale lookup operation to ground language using image search. |
InferLite: Simple Universal Sentence Representations from Natural Language Inference Data (D18-1)
Copied to clipboard
| Challenge: | InferLite is a lightweight version of InferSent that does not use recurrent layers and can generalize to multiple pre-trained word embeddings. |
| Approach: | They propose a lightweight version of InferSent that does not use recurrent layers and operates on a collection of pre-trained word embeddings. |
| Outcome: | The proposed model outperforms existing models that learn generic embeddings in an unsupervised setting, often requiring several days or weeks to train. |
Multichannel Generative Language Model: Learning All Possible Factorizations Within and Across Channels (2020.findings-emnlp)
Copied to clipboard
| Challenge: | MGLM is a generative joint distribution model over channels. |
| Approach: | They propose a multichannel generative joint distribution model over channels that marginalizes over all possible factorizations within and across all channels. |
| Outcome: | The proposed model outperforms traditional bilingual discriminative models. |
Character-Aware Models Improve Visual Text Rendering (2023.acl-long)
Copied to clipboard
Rosanne Liu, Dan Garrette, Chitwan Saharia, William Chan, Adam Roberts, Sharan Narang, Irina Blok, Rj Mical, Mohammad Norouzi, Noah Constant
| Challenge: | Current image generation models struggle to produce well-formed visual text due to lack of character-level input features. |
| Approach: | They conduct a series of experiments to compare character-aware vs. character-blind text encoders to determine their spelling ability. |
| Outcome: | The character-aware models outperform character-blind models on a range of novel text rendering tasks. |