computer vision papers have the best figures. this is a fact of life.
@jxmnop
-
Pretraining: The Beautiful Art of Learning to Compress Knowledge
By
–
pretraining, the act of learning-to-compress the entirety of human knowledge and thereby creating a general-purpose model that can simulate any natural process, is really a beautiful learning paradigm. everything is just going to get messier & more complicated from here
-
Three AI Researchers to Follow for Architecture Innovation
By
–
here are three awesome researchers everyone should follow: – Songlin (
@SonglinYang4 / phd at MIT),
– Will @lambdaviking (phd at NYU / gonna be a prof soon)
– Rulin (
@RulinShao / phd at UW)! and here is why:
1. if i had to bet on one person to develop an architecture that -
Membership Inference Scaling Laws: Empirical Evidence in ML Research
By
–
yeah. in particular, the membership inference scaling law is just empirical evidence for stuff that ppl like like @pratyushmaini already knew
-
Navigating Academia’s Gatekeepers in AI Research
By
–
i angered one of academia’s Final Bosses someone who will snidely post how you rediscovered their results, without actually reading your paper there will always be reviewer 2s around to take the fun out of things, to tell you your ideas dont matter. cite them and move on!
-
Functions for Loading, Storing, and Manipulating Information Systems
By
–
i guess the others are used to represent functions which load, store, and otherwise manipulate information
-
LLM Training: Fixed Capacity Model Reduces Per-Sample Memorization
By
–
hard to say since we mostly study the average case not the worst case. but I think “filling a fixed capacity” is a useful model for LLM learning, which implies that training on more data will force models to memorize less per-sample
-
FP32 Headliner Figures Discussed in AI Research Paper
By
–
the headliner figs are fp32! but this is discussed in the paper…and the thread…
-
Information Sharing and Generalization as Understanding Mechanism
By
–
i think it's the information-sharing/generalization part that makes what we'd call understanding happen