ywen99 Profile Banner
Yuxin Wen Profile
Yuxin Wen

@ywen99

Followers
530
Following
945
Statuses
115

PhD student @umdcs @ml_umd advised by @tomgoldsteincs. Intern @GoogleAI. Prev. intern @Apple, @GoogleDeepMind, @Sony. Security and Privacy in Machine Learning.

Joined December 2021
Don't wanna be here? Send us removal request.
@ywen99
Yuxin Wen
21 hours
RT @SeanMcleish: Introducing the Gemstones💎. 22 models ranging from 50M to 2B parameters, spanning 11 widths and 18 depths trained for 350B…
0
24
0
@ywen99
Yuxin Wen
3 days
RT @jonasgeiping: Ok, so I can finally talk about this! We spent the last year (actually a bit longer) training an LLM with recurrent d…
0
181
0
@ywen99
Yuxin Wen
3 days
RT @tomgoldsteincs: New open source reasoning model! Huginn-3.5B reasons implicitly in latent space 🧠 Unlike O1 and R1, latent reasoning…
0
273
0
@ywen99
Yuxin Wen
3 days
RT @iScienceLuvr: Scaling up Test-Time Compute with Latent Reasoning: A Recurrent Depth Approach We study a novel language model architect…
0
180
0
@ywen99
Yuxin Wen
29 days
@tonywu_71 @andimarafioti Great idea! We implemented a similar approach by compressing visual tokens into registers within the first three layers and discarding all visual tokens after these layers. You can find more details in our paper:
0
0
1
@ywen99
Yuxin Wen
2 months
RT @SeanMcleish: Why is addition hard for next token predictors? Come hear about our fix for this at #NeurIPS! We’re presenting Abacus Embe…
0
11
0
@ywen99
Yuxin Wen
2 months
RT @neeljain1717: Excited to present “Be Like a Goldfish: Don’t Memorize!” led by @ahans30 at NeurIPS, East Building #4709. In this work, w…
0
7
0
@ywen99
Yuxin Wen
2 months
RT @JiuhaiC: 🚨 New VLM Paper ! Florence-VL: Enhancing Vision-Language Models with Generative Vision Encoder and Depth-Breadth Fusion 1️⃣ A…
0
4
0
@ywen99
Yuxin Wen
3 months
RT @umiacs: A @UofMaryland team led by @bhatele & @tomgoldsteincs is a finalist for the prestigious ACM Gordon Bell Prize! Their framework…
0
5
0
@ywen99
Yuxin Wen
4 months
RT @alex_stein0: Excited to share our latest research on predicting events from historical data! This work achieves strong performance in e…
0
8
0
@ywen99
Yuxin Wen
4 months
RT @micahgoldblum: 📢I’ll be admitting multiple PhD students this winter to Columbia University 🏙️ in the most exciting city in the world!…
0
150
0
@ywen99
Yuxin Wen
4 months
RT @RawalRuchit: 1/ 🚀 Meet CinePile 2.0 – now with an adversarial pipeline to refine dataset quality and fresh evaluations on the latest Vi…
0
7
0
@ywen99
Yuxin Wen
4 months
RT @gowthami_s: 📃📣 No dataset is perfect, it is an iterative process to get it right! Introducing CinePile 2.0, an improved version of the…
0
10
0
@ywen99
Yuxin Wen
4 months
RT @haozhean36: "Call me by your name and I'll call you by mine." 🌈👩‍❤️‍👨👨‍❤️‍👨 Excited to share our new paper: "On the Influence of Gender…
0
6
0
@ywen99
Yuxin Wen
4 months
RT @m2saxon: "Bring Your Own Data!" From @k_saifullaah @neeljain1717 et al. Evaluating knowledge/biases in LMs with: - No labeled eval da…
0
6
0
@ywen99
Yuxin Wen
4 months
RT @neeljain1717: Soon at poster #10 👀👀👀
Tweet media one
0
6
0
@ywen99
Yuxin Wen
4 months
RT @neeljain1717: On the Amtrak heading over to #COLM2024, @k_saifullaah and I will present Bring Your Own Data! Self-Sensitivity Evaluatio…
0
8
0
@ywen99
Yuxin Wen
5 months
📢Think removing watermarks is easy? Maybe not this time! Join our NeurIPS competition for a chance to win from a $6,000 prize pool and test your methods at removing our carefully prepared watermarks!
@furongh
Furong Huang
5 months
🚨 Join the Challenge! 🚨 We're excited to announce the NeurIPS competition "Erasing the Invisible: A Stress-Test Challenge for Image Watermarks" running from September 16 to November 5. This is your chance to test your skills in a cutting-edge domain and win a share of our $6000 prize pool! 🔍 Competition Overview: ▶️ Tracks: Black-box Attacks & Beige-box Attacks ▶️ Goal: Remove invisible watermarks while preserving image quality. ▶️ Inspired by: WAVES Benchmark 🔗 Important Dates: ▶️ Sep 16 - Nov 5: Submission phase ▶️ Nov 5: Registration and submission close ▶️ Nov 20: Winning team announcement 🌐 Details and Registration: ▶️ Website: ▶️ Hosted on Codabench: ⏩ Beige-Box Track: ⏩Black-Box Track: 💡 Why Participate? ▶️ Validate the robustness of image watermarks under varying conditions. ▶️ Compete in a domain inspired by real-world challenges. ▶️ Collaborate with a global community of researchers and practitioners. 💰 Prize Pool: $6000 (and counting as more sponsors join us!) ▶️ Interested in sponsoring? Contact us at erasinginvisible@googlegroups.com or furongh@umd.edu. 🔨 A Personal Note: Hosting this competition has been a challenging yet rewarding experience. Working with a dedicated organizing team, developers, and legal experts across multiple parties, I've learned a lot. We can't wait to see what you bring to the table! A big shout out to our organizing team @mucongding @TahseenRab74917 @bang_an_ @chenghao_deng @SOURADIPCHAKR18 Anirudh Satheesh @MerhdadS @ywen99 Kyle Sang @Aakriti0503 @xuandongzhao Mo Zhou @anniehartley_ @lileics @yuxiangw_cs @vishalm_patel @FeiziSoheil @tomgoldsteincs @furongh 💐 and our sponsor, center for machine learning at UMD @ml_umd @umiacs @tomgoldsteincs! Don't miss out—register today and start preparing your submissions!
Tweet media one
0
0
13
@ywen99
Yuxin Wen
6 months
RT @A_v_i__S: 1/n The jailbreaking space is full of nifty discrete optimization techniques for crafting prompts. Of those PEZ is often used…
0
9
0
@ywen99
Yuxin Wen
6 months
RT @PandaAshwinee: Excited to share Lottery Ticket Adaptation (LoTA)! We propose a sparse adaptation method that finetunes only a sparse su…
0
46
0