Bing Yan
@bingyan4science
Followers
96
Following
158
Media
2
Statuses
16
AI+Chemistry PhD student @nyuniversity | Prev PhD of Chemistry @mit
New York
Joined September 2017
RT @yuntiandeng: Can we build an operating system entirely powered by neural networks?. Introducing NeuralOS: towards a generative OS that….
0
39
0
RT @RickyTQChen: We've open sourced Adjoint Sampling!. It's part of a bundled release showcasing FAIR's research and open source commitment….
github.com
code for "Adjoint Sampling: Highly Scalable Diffusion Samplers via Adjoint Matching" - facebookresearch/adjoint_sampling
0
23
0
RT @aaronjhavens: New paper out with FAIR(+FAIR-Chemistry):. Adjoint Sampling: Highly Scalable Diffusion Samplers via Adjoint Matching. We….
0
19
0
RT @NYUDataScience: @YuntianDeng from @UWaterloo presented at CILVR on internalizing reasoning in language models. His team's finetuning ap….
0
10
0
RT @yuntiandeng: How many reasoning tokens does OpenAI o1 use? It turns out they are almost always multiples of 64 (99+% of the time in 100….
0
47
0
RT @OpenCatalyst: Introducing Meta’s Open Materials 2024 (OMat24) Dataset and Models! All under permissive open licenses for commercial and….
0
174
0
RT @yuntiandeng: What do people use ChatGPT for? We built WildVis, an interactive tool to visualize the embeddings of million-scale chat da….
0
28
0
4/ CatScore is inspired by QAGS @W4ngatang @kchonyc @ml_perception and BartScore @WeizheY @gneubig @stefan_fee. CatScore adds instance-level evaluation to round-trip accuracy @pschwllr which focuses on system-level. Read the full CatScore paper here:
lnkd.in
This link will take you to a page that’s not on LinkedIn
0
0
4
3/ CatScore matches the discriminative power of quantum computations, w/ a Spearman ρ=0.85. CatScore is also superior at picking effective catalysts among thousands of candidates, w/ a mean reciprocal rank one order of magnitude higher than quantum computations.
1
0
2
2/ CatScore is orders of magnitude faster than traditional evaluation methods. Taking a typical Rh-based catalyst used in the asymmetric hydrogenation of olefins as an example, CatScore only takes 3 CPU seconds while the quantum computation takes 75 CPU hours.
1
0
2
1/ How to evaluate automatically designed catalysts? Challenge: multiple catalysts work well but usually only one ground truth. CatScore: use a reverse prediction model to compare catalysts' predicted product to target product. w/ @kchonyc now out at @digital_rsc!
2
5
10
RT @yuntiandeng: I am hiring NLP/ML PhD students at UWaterloo, home to 5 NLP professors! Apply by Dec 1. Strong consideration will be given….
0
130
0
RT @yuntiandeng: Can LMs solve reasoning tasks without showing their work? "Implicit Chain of Thought Reasoning via Knowledge Distillation"….
0
91
0
RT @yuntiandeng: Excited to share that I'm joining @UWCheritonCS as an Assistant Professor and @VectorInst as a Faculty Affiliate in Fall '….
0
19
0
RT @yuntiandeng: Meet the Chalk-Diagrams plugin for ChatGPT! Based on @danoneata & @srush_nlp's lib, it lets you create vector graphics wit….
0
1
0
RT @yuntiandeng: Ever wondered how nondeterministic GPT-4 is even with greedy decoding (T=0)? I built a website that asks GPT-4 to draw a u….
0
40
0