Zobrazeno 1 - 10
of 34
pro vyhledávání: '"Cui, Wanyun"'
Autor:
Cui, Wanyun, Wang, Qianle
This paper reveals the phenomenon of parameter heterogeneity in large language models (LLMs). We find that a small subset of "cherry" parameters exhibit a disproportionately large influence on model performance, while the vast majority of parameters
Externí odkaz:
http://arxiv.org/abs/2404.02837
AI-generated text has proliferated across various online platforms, offering both transformative prospects and posing significant risks related to misinformation and manipulation. Addressing these challenges, this paper introduces SAID (Social media
Externí odkaz:
http://arxiv.org/abs/2310.08240
Autor:
Cui, Wanyun, Wang, Qianle
Instructions augmentation is a crucial step for unleashing the full potential of large language models (LLMs) in downstream tasks. Existing Self-Instruct methods primarily simulate new instructions from a few initial instructions with in-context lear
Externí odkaz:
http://arxiv.org/abs/2310.04484
Autor:
Cai, Shuyang, Cui, Wanyun
ChatGPT brings revolutionary social value but also raises concerns about the misuse of AI-generated text. Consequently, an important question is how to detect whether texts are generated by ChatGPT or by human. Existing detectors are built upon the a
Externí odkaz:
http://arxiv.org/abs/2307.02599
Autor:
Cui, Wanyun, Chen, Xingran
Previous research has demonstrated that natural language explanations provide valuable inductive biases that guide models, thereby improving the generalization ability and data efficiency. In this paper, we undertake a systematic examination of the e
Externí odkaz:
http://arxiv.org/abs/2305.15520
Autor:
Cui, Wanyun, Chen, Xingran
Recent years have witnessed the emergence of textual commonsense knowledge bases, aimed at providing more nuanced and context-rich knowledge. The integration of external commonsense into language models has been shown to be a key enabler in advancing
Externí odkaz:
http://arxiv.org/abs/2305.15516
Autor:
Cui, Wanyun, Chen, Xingran
Publikováno v:
NeurIPS 2022
In this paper, we propose a new method for knowledge base completion (KBC): instance-based learning (IBL). For example, to answer (Jill Biden, lived city,? ), instead of going directly to Washington D.C., our goal is to find Joe Biden, who has the sa
Externí odkaz:
http://arxiv.org/abs/2211.06807
Autor:
Cui, Wanyun, Chen, Xingran
Rules have a number of desirable properties. It is easy to understand, infer new knowledge, and communicate with other inference systems. One weakness of the previous rule induction systems is that they only find rules within a knowledge base (KB) an
Externí odkaz:
http://arxiv.org/abs/2110.13577
Autor:
Cui, Wanyun, Chen, Xingran
We study how to enhance text representation via textual commonsense. We point out that commonsense has the nature of domain discrepancy. Namely, commonsense has different data formats and is domain-independent from the downstream task. This nature br
Externí odkaz:
http://arxiv.org/abs/2109.02572
Autor:
Cui, Wanyun, Yan, Sen
Knowledge distillation uses both real hard labels and soft labels predicted by teacher models as supervision. Intuitively, we expect the soft labels and hard labels to be concordant w.r.t. their orders of probabilities. However, we found critical ord
Externí odkaz:
http://arxiv.org/abs/2107.01412