Papers by Xiang Lisa Li
TempLM: Distilling Language Models into Template-Based Generators (2023.findings-acl)
Copied to clipboard
| Challenge: | Pretrained language models (PLMs) have greatly improved text generation, but they have also been known to produce unfaithful or inappropriate content. |
| Approach: | They propose a pretrained language model that is a template-based generator and uses it to generate a text. |
| Outcome: | The proposed model is more faithful than the original PLM and more fluent than prior template systems. |
Prefix-Tuning: Optimizing Continuous Prompts for Generation (2021.acl-long)
Copied to clipboard
| Challenge: | Fine-tuning is the prevalent paradigm for using large pretrained language models for downstream tasks, but it requires updating and storing all the parameters of the LM. |
| Approach: | They propose a lightweight alternative to fine-tuning for natural language generation tasks that optimizes a sequence of continuous vectors, which they call the prefix. |
| Outcome: | The proposed approach outperforms fine-tuning in the full data setting and extrapolates better to examples with topics that are unseen during training. |
Contrastive Decoding: Open-ended Text Generation as Optimization (2023.acl-long)
Copied to clipboard
Xiang Lisa Li, Ari Holtzman, Daniel Fried, Percy Liang, Jason Eisner, Tatsunori Hashimoto, Luke Zettlemoyer, Mike Lewis
| Challenge: | Using a language model, maximum probability is a poor decoding objective because it produces short and repetitive text. |
| Approach: | They propose a reliable decoding approach that optimizes a contrastive objective subject to a plausibility constraint. |
| Outcome: | The proposed approach outperforms four strong decoding algorithms in automatic and human evaluations across wikipedia, news and story domains. |
s1: Simple test-time scaling (2025.emnlp-main)
Copied to clipboard
Niklas Muennighoff, Zitong Yang, Weijia Shi, Xiang Lisa Li, Li Fei-Fei, Hannaneh Hajishirzi, Luke Zettlemoyer, Percy Liang, Emmanuel Candes, Tatsunori Hashimoto
| Challenge: | OpenAI’s o1 model showed this capability but did not publicly share its methodology, leading to many replication efforts. |
| Approach: | They curate a small dataset s1K with 1,000 reasoning questions based on three criteria we validate through ablations: difficulty, diversity, and quality. |
| Outcome: | The proposed model exceeds o1-preview on competition math questions by up to 27% (MATH and AIME24). |
Specializing Word Embeddings (for Parsing) by Information Bottleneck (D19-1)
Copied to clipboard
| Challenge: | Pre-trained word embeddings contain rich syntactic and semantic information. |
| Approach: | They propose a variational information bottleneck method to nonlinearly compress word embeddings . they compress each word embedded to either a discrete tag or a continuous vector . |
| Outcome: | The proposed method can be extended to other syntactic or semantic tasks. |
Posterior Control of Blackbox Generation (2020.acl-main)
Copied to clipboard
| Challenge: | Existing methods for conditional natural language generation are limited in their ability to produce controlled output. |
| Approach: | They propose to augment neural generation models with discrete control states learned through a structured latent-variable approach. |
| Outcome: | The proposed approach improves over benchmarks while providing fine-grained control. |