NeurIPS
The Conference on Neural Information Processing Systems, a leading academic venue in artificial intelligence and machine learning. In 2024, the conference received more than 15,000 paper submissions, reflecting its central role in the dissemination of research across the field.
Core Facts
DDPM (Denoising Diffusion Probabilistic Models) laid the theoretical foundation for diffusion models in 2020.
In 2014, the NeurIPS program committee conducted a review consistency experiment in which approximately 10% (about 166 papers) were reviewed by two independent program committees, and the two groups' acceptance decisions were only approximately 57% consistent.
Google's 2015 paper "Hidden Technical Debt in Machine Learning Systems" revealed that the actual machine learning code in ML systems accounts for only a very small proportion.
Timeline (last 90 days)
A senior reviewer has reviewed for top conferences such as NeurIPS, ICML, AAAI, and ICLR about 10 times since 2022.
NeurIPS and ICML in machine learning, SOSP and OSDI in systems, and CHI in human-computer interaction all belong to A*-class conferences.
The presence of undergraduate sole authors at A*-class conferences such as NeurIPS, ICML, SOSP, OSDI, and CHI is indeed increasing.
According to statistics in recent years, PyTorch accounts for more than 60% of paper implementations at conferences such as CVPR, NeurIPS and ICLR, significantly ahead of TensorFlow and JAX.
NeurIPS is one of the largest top conferences in the machine learning field, and in recent years its submissions have exceeded twenty thousand.
Top conferences such as NeurIPS and ICML likewise face an exponential growth trend in submission volume.
OpenReview is an open peer review platform widely adopted by current top AI conferences; ICLR has used it fully since 2018, and conferences such as NeurIPS and ICML have gradually followed suit.
Submission volumes at top AI conferences have experienced explosive growth over the past few years, with the number of submissions to conferences such as ICLR, NeurIPS and ICML expanding from a few thousand to tens of thousands, while the number of reviewers has not grown in tandem.
The official language of NeurIPS and ICML is English, with submissions, reviews and presentations all conducted in English, and their venues are mostly concentrated in North America and Europe.
NeurIPS was founded in 1987 and attracts more than 20,000 participants each year, with its paper acceptance rate long maintained at around 20% to 25%.
40 more timeline events
All Facts (20)
DDPM (Denoising Diffusion Probabilistic Models) laid the theoretical foundation for diffusion models in 2020.
90%VerifiedIn 2014, the NeurIPS program committee conducted a review consistency experiment in which approximately 10% (about 166 papers) were reviewed by two independent program committees, and the two groups' acceptance decisions were only approximately 57% consistent.
85%VerifiedGoogle's 2015 paper "Hidden Technical Debt in Machine Learning Systems" revealed that the actual machine learning code in ML systems accounts for only a very small proportion.
80%VerifiedICLR adopts an OpenReview mechanism, under which papers, review comments, author responses, and discussions are visible to the entire academic community during specific stages.
75%VerifiedThe NeurIPS 2019 conference introduced a reproducibility checklist requiring authors to provide code and experimental details.
75%UnverifiedThe NeurIPS paper acceptance rate is typically maintained between 20% and 25%.
70%UnverifiedNeurIPS is one of the most prestigious academic conferences in the field of machine learning.
70%UnverifiedNLP conferences typically use double-blind review, with each paper generally reviewed independently by 3–4 reviewers.
70%UnverifiedIn 2021, NeurIPS conducted a famous experiment in which the same paper was submitted twice and received completely different review outcomes, revealing the high degree of randomness in the review process.
70%UnverifiedThe results of the NeurIPS 2022 consistency experiment showed that approximately 50% of papers received completely different accept/reject decisions between the two groups of reviewers
70%UnverifiedNIPS was renamed NeurIPS in 2018
70%UnverifiedNeurIPS receives more than 10,000 submissions each year, with an acceptance rate typically around 25%.
70%UnverifiedThe 2020 DDPM (Denoising Diffusion Probabilistic Models) paper marked the point at which diffusion models began to surpass GANs.
70%UnverifiedNeurIPS 2023 received approximately 12,000 submissions, with approximately 3,500 ultimately accepted, yielding an acceptance rate of approximately 26%.
70%UnverifiedA senior reviewer has reviewed for top conferences such as NeurIPS, ICML, AAAI, and ICLR about 10 times since 2022.
50%UnverifiedThe presence of undergraduate sole authors at A*-class conferences such as NeurIPS, ICML, SOSP, OSDI, and CHI is indeed increasing.
50%UnverifiedNeurIPS and ICML in machine learning, SOSP and OSDI in systems, and CHI in human-computer interaction all belong to A*-class conferences.
50%UnverifiedAccording to statistics in recent years, PyTorch accounts for more than 60% of paper implementations at conferences such as CVPR, NeurIPS and ICLR, significantly ahead of TensorFlow and JAX.
50%UnverifiedNeurIPS is one of the largest top conferences in the machine learning field, and in recent years its submissions have exceeded twenty thousand.
50%UnverifiedTop conferences such as NeurIPS and ICML likewise face an exponential growth trend in submission volume.
50%