activity
20132026
most citedBeyond the Imitation Game: Quantifying and extrapolating the capabilities of language models

565 citations · 974 across the 50 of their papers we have counts for

collaborators
Showing 2023 · cs.CLShow all

6 papers · 2 filters

cs.CL2023

Do pretrained Transformers Learn In-Context by Gradient Descent?

Lingfeng Shen, Aayush Mishra, Daniel Khashabi

The emergence of In-Context Learning (ICL) in LLMs remains a remarkable phenomenon that is partially understood. To explain ICL, recent studies have created theoretical connections…

cs.CL2023★ 3 cited

SemStamp: A Semantic Watermark with Paraphrastic Robustness for Text Generation

Abe Bohan Hou, Jingyu Zhang, Tianxing He +7

Existing watermarking algorithms are vulnerable to paraphrase attacks because of their token-level design. To address this issue, we propose SemStamp, a robust sentence-level seman…

cs.CL2023

Error Norm Truncation: Robust Training in the Presence of Data Noise for Text Generation Models

Tianjian Li, Haoran Xu, Philipp Koehn +2

Text generation models are notoriously vulnerable to errors in the training data. With the wide-spread availability of massive amounts of web-crawled data becoming more commonplace…

cs.CL2023★ 1 cited

The Trickle-down Impact of Reward (In-)consistency on RLHF

Lingfeng Shen, Sihao Chen, Linfeng Song +5

Standard practice within Reinforcement Learning from Human Feedback (RLHF) involves optimizing against a Reward Model (RM), which itself is trained to reflect human preferences for…

cs.CL2023★ 6 cited

"According to ...": Prompting Language Models Improves Quoting from Pre-Training Data

Orion Weller, Marc Marone, Nathaniel Weir +3

Large Language Models (LLMs) may hallucinate and generate fake information, despite pre-training on factual data. Inspired by the journalistic device of "according to sources", we…

cs.CL2023

Flatness-Aware Prompt Selection Improves Accuracy and Sample Efficiency

Lingfeng Shen, Weiting Tan, Boyuan Zheng +1

With growing capabilities of large language models, prompting them has become the dominant way to access them. This has motivated the development of strategies for automatically se…