
Virginia Smith
@gingsmith
Followers
698
Following
12
Media
0
Statuses
38
RT @prthaker_: I'm very excited to share some new work This work started out in conversations with @thorn where we….
0
8
0
RT @icmlconf: ICML offers an optional poster printing service Orders can be picked up the day at the Vancouver Con….
0
3
0
RT @matthewyryang: 🚨 NEW PAPER: What if LLMs could tackle harder problems - not by explicitly training on longer traces, but by learning ho….
0
2
0
RT @mlcmublog: Are your LLMs truly forgetting unwanted data? . In this new blog post authored by @shengyuan_26734,….
0
3
0
RT @atalwalkar: I’m excited to share new work from Datadog AI Research! We just released Toto, a new SOTA (by a wide margin!) time series f….
0
54
0
RT @icmlconf: Invited talked are announced. Jon Kleinberg.Pamela Samuelson.Frauke Kreuter.Anca Dragan.Andreas Kr….
0
5
0
RT @AashiqMuhamed: Thrilled to share our new work on improving LLM unlearning! 🚀 .Gradient-based unlearning struggle with high cost, instab….
0
10
0
RT @mlcmublog: 📈⚠️ Is your LLM unlearning benchmark measuring what you think it is? . In a new blog post authored….
0
12
0
RT @andrewgwils: The ICML 2025 workshops list is online! Many exciting topics, spanning multi-agent systems, world….
0
17
0
RT @setlur_amrith: How to effectively unlearn finetuning data?.❌ Approx. methods leak sensitive data.✅ Exact unlearning (eg. retraining) is….
0
10
0
RT @icmlconf: ICML submission site is now open. Note that there's an abstract deadline on Jan 23.
0
24
0
RT @mlcmublog: How can we train LLMs to solve complex challenges beyond just data scaling? In a new blogpost, @set….
0
22
0
#4: Reviewing takes a lot of work and reviews themselves contain highly valuable information. In line with NeurIPS & ICLR, which have made accepted papers’ reviews public for years, reviews and discussions of all *accepted* ICML papers will be made public after the conference.
0
2
34
#3: The use of GenAI tools for reviewing is prohibited, for two key reasons. First, allowing AI generated reviews can damage the credibility of reviewing. Second, submissions and reviews may contain privileged information that should not be inputted into such tools.
2
1
16