Josh Robinson @_josh_robinson

Ph.D. student at @CSatUSC/@nlp_usc. Research: sequence modeling, #NLProc, #ML. joshua-robinson.github.io Los Angeles, CA Joined June 2018

Tweets

8
Followers

111
Following

347
Likes

3K

Michael Saxon @m2saxon

a year ago

@sintelion @BenZhou96 @muhao_chen Awesome analysis of what KNN-LM says abt training: Is the seeming "free lunch" of KNN-LM (replacing top LM layers with embedding store and KNN lookup) due to a weakness of the LM objctve? Seems no! Training a replacement MLP on the KNN does better! 🤔 aclanthology.org/2024.naacl-sho…

1 2 3 2K 3

Download Image

Ting-Rui Chiang @ctinray

a year ago

“On Retrieval Augmentation and the Limitations of Language Model Training” (arxiv.org/abs/2311.09615) has been accepted to NAACL 2024! While it is well known that kNN retrieval can decrease LMs’ perplexity, the underlying reason is unclear. We study two hypotheses 👇

1 6 15 4K 4

Aran Komatsuzaki @arankomatsuzaki

3 years ago

Leveraging Large Language Models for Multiple Choice Question Answering Finds that code/text-davinci performs much better on MCQ if the candidate answers are characters like "A", "B", etc unlike the original GPT3. arxiv.org/abs/2210.12353

2 10 46 0 16

Download Image

Ethan Perez @EthanJPerez

4 years ago

@rajammanabrolu @douwekiela @kchonyc Here are two I found quite interesting: 1. arxiv.org/abs/2111.13440 from @timo_schick @HinrichSchuetze 2. arxiv.org/abs/2203.11364 from @ChrisRytting et al.

0 2 8 0 1

Taylor Sorensen @ma_tay_

4 years ago

Most methods for aligning LMs to tasks require many labeled data (prompt eng) or access to the model (soft-prompts). In our work @ACL, we select high-perf. prompts 𝘸𝘪𝘵𝘩𝘰𝘶𝘵 𝘥𝘪𝘳𝘦𝘤𝘵 𝘢𝘤𝘤𝘦𝘴𝘴 𝘵𝘰 𝘵𝘩𝘦 𝘮𝘰𝘥𝘦𝘭 and 𝘸𝘪𝘵𝘩𝘰𝘶𝘵 𝘭𝘢𝘣𝘦𝘭𝘦𝘥 𝘦𝘹𝘢𝘮𝘱𝘭𝘦𝘴👇

1 1 4 0 0

Download Image

BYU @BYU

6 years ago

What started out as an opportunity for extra credit in one of his BYU classes led to a month-long, all-expenses-paid trip to China, a “pretty sweet trophy and a very good scholarship" for BYU senior Josh Robinson: news.byu.edu/intellect/byu-…