<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/"><channel><title>Nlp on Jiaqi Liu</title><link>https://jiaqiliu.blog/tags/nlp/</link><description>Recent content in Nlp on Jiaqi Liu</description><generator>Hugo</generator><language>en-US</language><lastBuildDate>Tue, 28 Oct 2025 12:00:00 -0700</lastBuildDate><atom:link href="https://jiaqiliu.blog/tags/nlp/index.xml" rel="self" type="application/rss+xml"/><item><title>MemeBLIP2: A Lightweight Multimodal System to Detect Harmful Memes</title><link>https://jiaqiliu.blog/posts/memeblip2/</link><pubDate>Tue, 28 Oct 2025 12:00:00 -0700</pubDate><guid>https://jiaqiliu.blog/posts/memeblip2/</guid><description>A lightweight multimodal system that pairs the BLIP-2 encoder with joint image-text modelling to detect harmful memes (IJCAI 2025 workshop paper).</description></item><item><title>Lightweight Baselines for Medical Abstract Classification</title><link>https://jiaqiliu.blog/posts/medical-abstract-classification/</link><pubDate>Sun, 19 Oct 2025 00:00:00 +0000</pubDate><guid>https://jiaqiliu.blog/posts/medical-abstract-classification/</guid><description>Code and results behind our MLNLP 2025 paper: DistilBERT with plain cross-entropy is a strong default for classifying medical abstracts.</description></item></channel></rss>