Visit the post for more.  ( 8 min )

Is superforecasting just sneaky accounting?

Is superforecasting just sneaky accounting?

Submitted by Charles Belov: I've been browsing through the proposed Unicode 17 changes, currently undergoing a comment period, with interest. While I don't have the knowledge to intelligently comment on the proposals, it's good to see that they are actively improving language access. I'm puzzled that some new characters have been added to the existing […]  ( 11 min )

Machine learning workflows often involve a delicate balance: you want models that perform exceptionally well, but you also need to understand and explain their predictions.

Visit the post for more.  ( 8 min )

We've talked about Dungan a lot on Language Log.  That's the northwest Sinitic topolect written in Cyrillic that has been transplanted to Central Asia.  See "Selected readings" below. For those of you who are interested and would like to hear what it sounds like in real life — spoken and sung by male and female […]  ( 11 min )

This post is divided into five parts; they are: • Understanding Positional Encodings • Sinusoidal Positional Encodings • Learned Positional Encodings • Rotary Positional Encodings (RoPE) • Relative Positional Encodings Consider these two sentences: "The fox jumps over the dog" and "The dog jumps over the fox".

Visit the post for more.  ( 8 min )

Running down the road in Clarksdale, Mississippi, I screeched to a halt (felt like Rroad Runner) when I passed by a Chinese restaurant with the odd name Rice Bowl (in Chinese it was Fànwǎn lóu 饭碗楼 — the only characters I saw on the premises).  It was a tiny, nondescript establishment, with six or so […]  ( 17 min )

Visit the post for more.  ( 8 min )

The field of large language models is shifting toward lower-precision computation. This shift necessitates a rethinking of scaling laws to account for the effects of quantization on resulting quantized model...  ( 42 min )

That's the title of a valuable Wikipedia article.  I have no idea who wrote it, but I'm very glad to have access to this comprehensive article, since it touches on so many topics that concern my ongoing research. Here are some highlights: Before British colonisation, the Persian language was the lingua franca of the Indian […]  ( 13 min )

The first two panels from SMBC a few days ago: The rest of the strip: The aftercomic: The mouseover title: "Would you rather sit with friends watching shadows on the bigscreen or spending your time arguing with Plato about whether poetry should be legal?" This expands on the 9/9/2015 SMBC: Wikipedia explains the Allegory of […]  ( 9 min )

"Does GPT-4 Surpass Human Performance in Linguistic Pragmatics?" Bojic, Ljubiša et al. Humanities and Social Sciences Communications 12, no. 1 (June 10, 2025). Ljubiša Bojić, Predrag Kovačević, & Milan Čabarkapa.  Humanities and Social Sciences Communications volume 12, Article number: 794 (2025) Cite this article Abstract As Large Language Models (LLMs) become increasingly integrated into everyday life as general-purpose […]  ( 10 min )

Validity as a style guide for telling stories about correlations

Validity as a style guide for telling stories about correlations

We partnered with Darren Aronofsky, Eliza McNitt and a team of more than 200 people to make a film using Veo and live-action filmmaking.  ( 16 min )

Pandas , NumPy , and Scikit-learn .

这篇文章我们来推导$\newcommand{msign}{\mathop{\text{msign}}}\msign$算子的求导公式。如果读者想要像《Test-Time Training Done...  ( 6 min )

From Adam Rasgon and Natan Odenheimer, "U.S. Embassy in Jerusalem Braces for Possible Israeli Strike on Iran" NYT 6/12/2025: More recently, however, Mr. Trump has said he was less convinced that talks with Iran would yield a new nuclear deal. “I’m getting more and more less confident about it,” he told The New York Post […]  ( 10 min )

In a comment on yesterday's post "A 12th-century influencer", Laura Morland wrote: Thanks for sharing "to abelard," the new verb of the month! Note to AP: the grammarians will insist that it be spelled with a lower-case "a". (Verbs are never capitalized, not even in German, I don't believe.) This is one where The Errorist […]  ( 12 min )

The first two panels of today's xkcd: The rest of it: The mouseover title: "If you think curiosity without rigor is bad, you should see rigor without curiosity." It's not just science — today's Tank McNamara: Some extra reading: Gavin Francis, "What Do You Expect?", The New York Review 6/26/2025. A couple of relevant past […]  ( 11 min )

From Ting Fen Yik on Facebook: It's been a while since we've posted on ambigrams.  David Moser is the master in Chinese and in English.  See the references below.   Selected readings "Weird characters" (7/7/13) "Orientation-dependent ambiguity" (12/27/18) "Sinographic memory in Vietnamese writing" (4/16/14) — see esp. the last comment "Freemocracy" (6/13/19) "Happy LÓNG year!" […]  ( 9 min )

AI Narratives: A history of imaginative thinking about intelligent machines is a 2020 collective book edited by Stephen Cave, Kanta Dihal, and Sarah Dillon, with about twenty contributing authors, through a series of 16 chapters on the relation between culture (literature, films) and our societal approach to AI, with varying perspectives, some overlap between chapters, […]  ( 10 min )

We’re launching Weather Lab, featuring our experimental cyclone predictions, and we’re partnering with the U.S. National Hurricane Center to support their forecasts and warnings this cyclone season.  ( 7 min )

The Trevisan Prize for outstanding work in the Theory of Computing is sponsored by the Department of Computing Sciences at Bocconi University and the Italian Academy of Sciences. The prize is named in honor of Luca Trevisan in recognition of his major contributions to the Theory of Computing. It aims to recognize outstanding work in the field, and to broaden the reach … Continue reading Trevisan prize (guest post by Alon Rosen)  ( 13 min )

The Trevisan Prize for outstanding work in the Theory of Computing is sponsored by the Department of Computing Sciences at Bocconi University and the Italian Academy of Sciences. The prize is named in honor of Luca Trevisan in recognition of his major contributions to the Theory of Computing. It aims to recognize outstanding work in the field, and to broaden the reach … Continue reading Trevisan prize (guest post by Alon Rosen)  ( 13 min )

Imbalanced datasets, where a majority of the data samples belong to one class and the remaining minority belong to others, are not that rare.

Read La Fille du Grand Hiver (The Daughter of the Great Winter) by Isabelle Autissier (also a sailor who was the first woman to complete a solo world race in 1991). This is a novelised version of the story of Arnarulunguaq, who accompanied Knud Rasmussen on the Fifth Thule Expedition over several years, all the […]  ( 10 min )

From Ada Palmer, "Inventing the Renaissance: The Myth of a Golden Age": The new scholastic method was so exciting! that when Peter Abelard got kicked out of his monastery (for proving its founding saint didn’t exist—that pissed off the abbot, who’d have guessed?) and went to live as a hermit in the wilderness of Champagne, […]  ( 11 min )

The latest xkcd: Mouseover title: "With a good battery, the device can easily last for 5 or 10 years, although the walls probably won't." The joke worked for me, although I was pretty sure that a (current) MacBook makes no sound when a usb device connects. I checked, and that's true. A current Windows 11 […]  ( 10 min )

Randy Alexander is not a professional Sinologist, but when it comes to reading Chinese poetry, he's as serious as one can be.  The following poem is by Du Fu (712-770), said by some to be "China's greatest poet".  In the presentation below, I will first give the text with its transcription, and then Randy's translation.  […]  ( 17 min )

You've trained your machine learning model, and it's performing great on test data.

无需蒸馏、抛弃SFT，Mistral用纯强化学习在数学与代码推理任务上实现50%性能飞跃 近日，Mistral […]  ( 4 min )

核心突破：极简激活的超级大脑 三大技术支柱撑起SOTA表现 1. 数据工程：11.2T高质量token的炼金术 […]  ( 4 min )

Trevisan Prize 2025 Here is a call for nominations for a new theoretical computer science prize, in memory of  Luca Trevisan. (h/t Alon Rosen.) Three Near Future Events at HUJI While the Erdős Lectures 2025 given by Mehtaab S. Sawhney, … Continue reading →  ( 14 min )

Trevisan Prize 2025 Here is a call for nominations for a new theoretical computer science prize, in memory of  Luca Trevisan. (h/t Alon Rosen.) Three Near Future Events at HUJI While the Erdős Lectures 2025 given by Mehtaab S. Sawhney, … Continue reading →  ( 14 min )

In this new presidential order of 23 May 2025, Trump pretends to “restore the scientific integrity policies of my first Administration and ensures that agencies practice data transparency, acknowledge relevant scientific uncertainties, are transparent about the assumptions and likelihood of scenarios used, approach scientific findings objectively, and communicate scientific data accurately” repeating his goal in […]  ( 10 min )

Remind me what happens when a measure becomes a target.

Remind me what happens when a measure becomes a target.

There’s no doubt that search is one of the most fundamental problems in computing.

Exceptionally, the next One World Approximate Bayesian Inference (OWABI) Seminar will be hybrid as it is scheduled to take place during BayesComp 2025 in Singapore, on Thursday 19 June at 8pm Singapore time (1pm in Tórshavn) and two talks, one by Filippo Pagani on Approximate Bayesian Fusion Bayesian Fusion is a powerful approach that enables […]  ( 10 min )

The rise of language models, and more specifically large language models (LLMs), has been of such a magnitude that it has permeated every aspect of modern AI applications — from chatbots and search engines to enterprise automation and coding assistants.

That's the title of a brand new (3/13/25) book by Laura Spinney, author of Pale Rider, a noteworthy volume on the 1918 influenza pandemic.  Here she is interviewed (6/7/25) by Colin Gorrie (the interview is too long [58:14] to post directly on Language Log): Proto-Indo-European Origins: A Conversation with Laura Spinney     Follow along with the […]  ( 9 min )

Babbel's April 2025 Semicolon Survey looked at students' reactions to the obvious secular decline in semicolon frequency: The semicolon once stood as a symbol of thoughtful, elegant writing, a punctuation mark beloved by literary greats like Jane Austen and Virginia Woolf. But today, the humble semicolon faces an uncertain future. New analysis from Babbel uncovers […]  ( 13 min )

A letter to the editor on the essential nature of LLMs from the Times Literary Supplement (5/30/25):  Large language models As someone who has spent the past few years working out what AI means to academic journals, I found Melanie Mitchell’s excellent review of These Strange New Minds by Christopher Summerfield (May 16) full of […]  ( 11 min )

Visit the post for more.  ( 9 min )

This 2023 paper of Perepolkin, Goodrich, and Sahlin vaguely relates to our insufficient Gibbs work in that a Bayesian analysis is conducted based solely on quantile summaries. Except that here the input is the entire cdf, or the—inverse cdf—quantile function, or—its derivative—the quantile density function, instead of the probability density function—used as the likelihood in […]  ( 10 min )

Stephanie Farr, "The nonbinary Revolutionary leader who preached in Philly during the Revolution", The Philadelphia Inquirer 6/5/2025: Sometimes when I walk the streets of Old City, I imagine the people of colonial times who walked those roads before me, before Philadelphia was Philly and before this nation secured its liberty and identity. I mostly think […]  ( 10 min )

Faizan Zaki overcomes a shocking, self-inflicted flub and wins the Scripps National Spelling Bee Ben Nuckols, AP (5/30/25) Not what you would expect when the stakes are so high: The favorite entering the bee after his runner-up finish last year — during which he never misspelled a word in a conventional spelling round, only to […]  ( 12 min )

Visit the post for more.  ( 9 min )

Summary  PyTorch Distributed Checkpointing (DCP) is making investments into addressing the interoperability blockers to ensure that popular formats, like HuggingFace safetensors, can work well with PyTorch’s ecosystem. Since HuggingFace has...  ( 38 min )

As most readers of Language Log know, ABC means "American-born Chinese".  Depending upon how (in)sensitive their parents are, learning Chinese can be hell, and leave them scarred for life. The actors in this video are brilliant and the tale it tells reveals so much about the trials and pitfalls of learning Chinese overseas. If only […]  ( 10 min )

The news has been full of the Musk-Trump feud. Among the linguistic aspects, there's an interesting amount of explicit or implied gender association — here's Alexandria Ocasio-Cortez in a memic clip widely linked on social media: Your browser does not support the video tag. From the other end of the political spectrum, check out Nellie […]  ( 11 min )

Following up on the issue of English spelling variation, this picture has been making the rounds on social media: I thought of it when I was reminded that the New Jersey borough of Buena is pronounced /ˈbjuːnə/ — so that the first syllable is the same as the first syllable of beauty. It's not clear […]  ( 13 min )

For English speakers, a mind-boggling letter to the editor on linguistic gender from the Times Literary Supplement (3/9/25): Masculine and feminine In Cristina Rivera Garza’s Death Takes Me, reviewed by Lucy Popescu (In Brief, April 18), a character points out that “in Spanish, the word victim, or victima, is always feminine”. This is evidently true, but […]  ( 12 min )

前面我们用了两篇文章《msign算子的Newton-Schulz迭代（上）》和《msign算子的Newton-Schulz迭代（下）》讨论了矩阵的$\newcommand{msign}{\mat...  ( 6 min )

Missing values appear more often than not in many real-world datasets.

阿里巴巴通义实验室发布的 Qwen3 Embedding 系列模型在文本嵌入（Embedding）和重排序（R […]  ( 4 min )

A recent arXival by Liviu Aolaritei, Bart Van Parys, Henry Lam, and Michael Jordan (a co-PI in our ERC Synergy Ocean project) discusses optimal importance sampling schemes for stochastic optimisation, processed by an iterative Robbins-Munro algorithm improvement (with the Polyak-Ruppert improvement). “Despite its popularity, IS is often described as a `double-edged sword.’ Its performance depends […]  ( 10 min )

The PyTorch Ecosystem goes back several years, with some of its earliest projects like Hugging Face, Fast.ai, and PyTorch Lightning going on to grow incredible communities of their own. The...  ( 41 min )

In Daniel Dennett's 1995 book Darwin's Dangerous Idea: Evolution and the Meanings of Life, the chapter titled "Chomsky contra Darwin, Four Episodes" ends with this provocative sentence: The hostility to Artificial Intelligence and its evil twin, Darwinism, lies just beneath the surface of much of the most influential work in recent twentieth-century philosophy. What Dennett […]  ( 14 min )

Elizabeth Ribbens, "How the use of a word in the Guardian has gotten some readers upset", The Guardian 6/4/2025: ‘Got’ was changed during the editing of an opinion piece, leading to correspondence lamenting a slide into American English. But language isn’t a fortress. In Shakespeare’s Henry VI, Part II, a messenger breathlessly announces to the […]  ( 13 min )

I must say, with the ongoing hype around machine learning, a lot of people jump straight to the application side without really understanding how things work behind the scenes.

Machine learning is not just about building models.

If peer review only refers to papers, it is not defensible

If peer review only refers to papers, it is not defensible

在上文《msign算子的Newton-Schulz迭代（上）》中，我们试图为$\mathop{\text{msign}}$算子寻找更好的Newton-Schulz迭代，以期在有限迭代步数内能达到...  ( 8 min )

(Click to enlarge) Today, at 14:30, Monday June 9 at 11:00 and Wednesday June 11, at 11:20 Mehtaab S. Sawhney will deliver the 2005 Erdos lecture. Three talks, each representing a monumental achievement!  ( 14 min )

(Click to enlarge) Today, at 14:30, Monday June 9 at 11:00 and Wednesday June 11, at 11:20 Mehtaab S. Sawhney will deliver the 2005 Erdos lecture. Three talks, each representing a monumental achievement!  ( 14 min )

Visit the post for more.  ( 9 min )

More than 20 years ago, I posted about the explosion of -ome and -omic words in biology: "-ome is where the heart is", 10/27/2004. I listed more than 40 examples: behaviourome, cellome, clinome, complexome, cryptome, crystallome, ctyome, degradome, enzymome,epigenome, epitome, expressome, fluxome, foldome, functome, glycome, immunome, ionome, interactome, kinome, ligandome, localizome, metallome, methylome, morphome, nucleome, […]  ( 11 min )

No content preview  ( 40 min )

The tools we use to build AI are evolving fast, with PyTorch at the heart of many advances. But unless we evolve the way we approach building AI systems, we...  ( 43 min )

Machine learning workflows typically involve plenty of numerical computations in the form of mathematical and algebraic operations upon data stored as large vectors, matrices, or even tensors — matrix counterparts with three or more dimensions.

Visit the post for more.  ( 9 min )

A brief collection of "Chinese words for Adults!", with the last one being "KPI", which I had to look up in English. Posted by UFL – University Of Foreign Languages – LE on Monday, May 26, 2025 A performance indicator or key performance indicator (KPI) is a type of performance measurement. KPIs evaluate the success of an organization or of a particular activity […]  ( 10 min )

Gemini 2.5 has new capabilities in AI-powered audio dialog and generation.  ( 14 min )

Wait, what?

Wait, what?

Ethereum Foundation talk, today This afternoon (Tuesday, June 3, 2025) at 17:00 Israel time I give a zoom lecture on A Critical View on Quantum Computing. The lecture is hosted by the Ethereum Foundation and the 90 minute events will … Continue reading →  ( 15 min )

Ethereum Foundation talk, today This afternoon (Tuesday, June 3, 2025) at 17:00 Israel time I give a zoom lecture on A Critical View on Quantum Computing. The lecture is hosted by the Ethereum Foundation and the 90 minute events will … Continue reading →  ( 15 min )

Feature engineering is a key process in most data analysis workflows, especially when constructing machine learning models.

Visit the post for more.  ( 9 min )

This post is divided into three parts; they are: • Understanding Word Embeddings • Using Pretrained Word Embeddings • Training Word2Vec with Gensim • Training Word2Vec with PyTorch • Embeddings in Transformer Models Word embeddings represent words as dense vectors in a continuous space, where semantically similar words are positioned close to each other.

最近在阅读时，遇到了一个关于最优多项式逼近的“等值振荡定理（Equioscillation Theorem）”，证明过程还涉及到无穷范数求导，感觉结论和证明都颇为新奇，特来记录一番。参考资料：《...  ( 7 min )

Visit the post for more.  ( 9 min )

My friend EJ Wagenmakers, along with his colleague  Raoul Grasman, have proposed a novel measure of discrepancy between two distributions that they formulate through a basic Bayesian lens as an expected posterior probability, written as (with no priors harmed in the process!) in case both distributions are absolutely continuous wrt a common dominating measure, with […]  ( 9 min )

Visit the post for more.  ( 9 min )

Machine learning models have become increasingly sophisticated, but this complexity often comes at the cost of interpretability.

I would like to draw your attention to the short beautiful paper Hypergraphic Zonotopes and Acyclohedra by Cosmin Pohoata and Daniel G. Zhu. The paper introduces higher-uniformity analogue of graphic zonotopes and permutohedra, and provides formulas for their volume, and, … Continue reading →  ( 15 min )

I would like to draw your attention to the short beautiful paper Hypergraphic Zonotopes and Acyclohedra by Cosmin Pohoata and Daniel G. Zhu. The paper introduces higher-uniformity analogue of graphic zonotopes and permutohedra, and provides formulas for their volume, and, … Continue reading →  ( 15 min )

近日，小米开源社区发布了MiMo-VL-7B视觉语言模型技术报告，其SFT（监督微调）和RL（强化学习）版本在 […]  ( 3 min )

Who should define what is "good" and "bad" science?

Who should define what is "good" and "bad" science?

Quantization is a frequently used strategy applied to production machine learning models, particularly large and complex ones, to make them lightweight by reducing the numerical precision of the model’s parameters (weights) — usually from 32-bit floating-point to lower representations like 8-bit integers.

This post is divided into five parts; they are: • Naive Tokenization • Stemming and Lemmatization • Byte-Pair Encoding (BPE) • WordPiece • SentencePiece and Unigram The simplest form of tokenization splits text into tokens based on whitespace.

Machine learning model development often feels like navigating a maze, exciting but filled with twists, dead ends, and time sinks.

As part of a day-long meeting on organization of science in Prague, I gave a 10-minite presentation on Science and Diversity in a small Country (click for my slides) and devoted a few minutes and one slide  to the amazing … Continue reading →  ( 14 min )

As part of a day-long meeting on organization of science in Prague, I gave a 10-minite presentation on Science and Diversity in a small Country (click for my slides) and devoted a few minutes and one slide  to the amazing … Continue reading →  ( 14 min )

On May 17, the PyTorch Meetup was successfully held in Hangzhou, drawing nearly 60 developers and industry experts from companies including Huawei, Tencent, Ant Group, and ByteDance. The event focused...  ( 40 min )

There is no reproducibility crisis in science. So what is JD Vance on about?

There is no reproducibility crisis in science. So what is JD Vance on about?

近年来，大型推理模型（Large Reasoning Models, LRMs）在数学、编程和逻辑推理等任务中 […]  ( 4 min )

With members of the discrete geometry seminar in Prague. The 2025 Colloquia in Combinatorics took place in QMUL and UCL, 7 and 8 May 2025. 2025 is the 18th year of the Colloquia in Combinatorics, and like every year, the … Continue reading →  ( 15 min )

With members of the discrete geometry seminar in Prague. The 2025 Colloquia in Combinatorics took place in QMUL and UCL, 7 and 8 May 2025. 2025 is the 18th year of the Colloquia in Combinatorics, and like every year, the … Continue reading →  ( 15 min )

众所周知，生成速度慢是扩散模型一直以来的痛点，而为了解决这个问题，大家可谓“八仙过海，各显神通”，提出了各式各样的解决方案，然而长久以来并没一项工作能够脱颖而出，成为标配。什么样的工作能够达到这...  ( 7 min )

Open source licensing revolutionized software development, creating a thriving ecosystem built on shared innovation and collaboration. Licenses like MIT and Apache-2.0 gave developers a standard, legally robust way to share...  ( 40 min )

Featured Sessions: Exploring Innovation at PyTorch Day China 2025 PyTorch Day China 2025, proudly hosted by the PyTorch Foundation, will take place on June 7 in Beijing, China collocated with...  ( 38 min )

A tutorial explaining some key intuitions behind continuous time Markov chains for machine learners interested in discrete diffusion models: alternative representations, connections to point processes, and the memoryless property.  ( 9 min )

You have a large JSON file, and you want to load the data into Pydantic. Unfortunately, this uses a lot of memory, to the point where large JSON files are very difficult to read. What to do? Assuming you’re stuck with JSON, in this article we’ll cover: The high memory usage you get with Pydantic’s default JSON loading. How to reduce memory usage by switching to another JSON library. Going further by switching to dataclasses with slots. Read more...  ( 4 min )

The PyTorch ATX Triton event, sponsored by Red Hat, was held on April 30, 2025, at the University of Texas. It was an exciting gathering focused on the Triton framework...  ( 34 min )

Later this evening on 18:00 Prague time there will be a debate between Matthias Christandl and me, about the possibility of quantum computation. The debate will be live-streamed on YouTube. The beautiful poster was designed by Helena Gráfová. Click here … Continue reading →  ( 19 min )

Later this evening on 18:00 Prague time there will be a debate between Matthias Christandl and me, about the possibility of quantum computation. The debate will be live-streamed on YouTube. The beautiful poster was designed by Helena Gráfová. Click here … Continue reading →  ( 19 min )

We’ve made Gemini 2.5 our most secure model family to date.  ( 6 min )

Introducing Veo 3 and Imagen 4, and a new tool for filmmaking called Flow.  ( 16 min )

We’re extending Gemini to become a world model that can make plans and imagine new experiences by simulating aspects of the world.  ( 15 min )

Learn about the new SynthID Detector portal we announced at I/O to help people understand how the content they see online was generated.  ( 15 min )

Gemma 3n is a cutting-edge open model designed for fast, multimodal AI on devices, featuring optimized performance, unique flexibility with a 2-in-1 model, and expanded multimodal understanding with audio, empowering developers to build live, interactive applications and sophisticated audio-centric experiences.  ( 5 min )

Gemini 2.5 Pro continues to be loved by developers as the best model for coding, and 2.5 Flash is getting even better with a new update. We’re bringing new capabilities to our models, including Deep Think, an experimental enhanced reasoning mode for 2.5 Pro.  ( 17 min )

Alberto Testolin  ( 2 min )

Alan Chung, Amin Saberi, Morgane Austern  ( 2 min )

Sujay Nagaraj, Walter Gerych, Sana Tonekaboni, Anna Goldenberg, Berk Ustun, Thomas Hartvigsen  ( 2 min )

Ayush Jain, Andrea Montanari, Eren Sasoglu  ( 2 min )

Alexander Kozachinskiy, Tomasz Steifer  ( 2 min )

Matthias C. Caro, Marcel Hinsche, Marios Ioannou, Alexander Nietner, Ryan Sweke  ( 3 min )

Makoto Yamada, Yuki Takezawa, Guillaume Houry, Kira Michaela Dusterwald, Deborah Sulem, Han Zhao, Yao-Hung Hubert Tsai  ( 3 min )

Rob Brekelmans, Frank Nielsen  ( 2 min )

Ruiquan Huang, Yingbin Liang, Jing Yang  ( 2 min )

Jia-Fong Yeh, Hsin-Ying Lee, Bing-Chen Tsai, Yi-Rong Chen and Ping-Chia Huang, Winston H. Hsu  ( 2 min )

Hiram H. Lopez, Rafael H. Villarreal  ( 2 min )

Cicero Carvalho, V. G. Lopez Neumann, Hiram H. Lopez  ( 2 min )

Hiram H. Lopez, Carlos Renteria, Rafael H. Villarreal  ( 2 min )

Hiram H. Lopez, Eliseo Sarmiento, Maria Vaz Pinto, Rafael H. Villarreal  ( 2 min )

Hiram H. Lopez, Rafael H. Villarreal  ( 2 min )

Hiram H. Lopez, Rafael H. Villarreal, Leticia Zarate  ( 2 min )

Javier de la Cruz, Elisa Gorla, Hiram H. Lopez, Alberto Ravagnani  ( 2 min )

Eduardo Hitomi, Felipe Yasumura  ( 2 min )

Mohammad Daher, Khalil Saadi  ( 2 min )

Apoloniusz Tyszka  ( 3 min )

E. Abdalla, Jeferson de Oliveira, A. B. Pavan, C. E. Pellicer  ( 2 min )

Jan Palczewski, Lukasz Stettner  ( 2 min )

Mohammad Reza Rahmati, Gerardo Flores  ( 3 min )

Hans Havlicek, Peter \v{S}emrl  ( 2 min )

Hans Havlicek  ( 2 min )

Andrea Blunck, Hans Havlicek  ( 2 min )

Andrea Blunck, Hans Havlicek, Corrado Zanella  ( 2 min )

Andrea Blunck, Hans Havlicek, Corrado Zanella  ( 2 min )

Hannah Blocher, Georg Schollmeyer, Malte Nalenz, Christoph Jansen  ( 2 min )

Theodore Papamarkou, Maria Skoularidou, Konstantina Palla, Laurence Aitchison, Julyan Arbel, David Dunson, Maurizio Filippone, Vincent Fortuin, Philipp Hennig, Aliaksandr Hubin, Alexander Immer, Theofanis Karaletsos, Mohammad Emtiyaz Khan, Agustinus Kristiadi, Yingzhen Li, Jose Miguel Hernandez Lobato, Stephan Mandt, Christopher Nemeth, Michael A. Osborne, Tim G. J. Rudner, David R\"ugamer, Yee Whye Teh, Max Welling, Andrew Gordon Wilson, Ruqi Zhang  ( 2 min )

Yue Xing, Xiaofeng Lin, Namjoon Suh, Qifan Song, Guang Cheng  ( 2 min )

Vikranth Dwaracherla, Seyed Mohammad Asghari, Botao Hao, Benjamin Van Roy  ( 2 min )

Susumu Oda  ( 3 min )

Nicolas B. Cowan, Yuka Fujii  ( 2 min )

Mubarak Ali  ( 3 min )

Rupert L. Frank, Dirk Hundertmark, Michal Jex, Phan Th\`anh Nam  ( 2 min )

Hsueh-Yung Lin  ( 2 min )

Hsueh-Yung Lin  ( 2 min )

James P. Gleeson, Tomokatsu Onaga, Peter Fennell, James Cotter, Raymond Burke, David J. P. O'Sullivan  ( 2 min )

Eugene Gorsky, Matthew Hogancamp, Anton Mellit  ( 2 min )

Wentao Zhan, Abhirup Datta  ( 2 min )

Shengchao Liu, Weili Nie, Chengpeng Wang, Jiarui Lu, Zhuoran Qiao, Ling Liu, Jian Tang, Chaowei Xiao, Anima Anandkumar  ( 2 min )

Shuhei Kashiwamura, Ayaka Sakata, Masaaki Imaizumi  ( 2 min )

Ronald Richman, Salvatore Scognamiglio  ( 2 min )

Shuo Shuo Liu  ( 2 min )

Binghui Peng  ( 2 min )

Daesoo Lee, Erlend Aune  ( 2 min )

Shengchao Liu, Weitao Du, Yanjing Li, Zhuoxinran Li, Vignesh Bhethanabotla, Nakul Rampal, Omar Yaghi, Christian Borgs, Anima Anandkumar, Hongyu Guo, Jennifer Chayes  ( 2 min )

Daniel Z\"ugner, Stephan G\"unnemann  ( 2 min )

Zhenyu He, Guhao Feng, Shengjie Luo, Kai Yang, Di He, Jingjing Xu, Zhi Zhang, Hongxia Yang, Liwei Wang  ( 2 min )

Lea Beneish, Debanjana Kundu, Anwesh Ray  ( 2 min )

Julian Holstein, Andrey Lazarev  ( 2 min )

Mohammad M. Jalalzai, Jianyu Niu, Chen Feng, Fangyu Gai  ( 2 min )

\'Italo Dowell, Sergio Roma\~na  ( 2 min )

Sourav Ghosh, Nicolaus Treib  ( 2 min )

Terry Lyons, Andrew D. McLeod  ( 3 min )

Yi-Lin Sung, Jaehong Yoon, Mohit Bansal  ( 3 min )

Augustina C. Amakor, Konstantin Sonntag, Sebastian Peitz  ( 3 min )

Vishwali Mhasawade, Salman Rahman, Zoe Haskell-Craig, Rumi Chunara  ( 2 min )

Boxin Zhao, Percy S. Zhai, Y. Samuel Wang, Mladen Kolar  ( 3 min )

Jurek Czyzowicz, Dariusz Dereniowski, Andrzej Pelc  ( 2 min )

Or Hershkovits, Brian White  ( 2 min )

Varghese Mathai, David Roberts  ( 2 min )

C. Baumgarten  ( 2 min )

Yoshinosuke Hirakawa  ( 2 min )

Andrei Tomut, Saeed S. Jahromi, Sukhbinder Singh, Faysal Ishtiaq, Cesar Mu\~noz, Prabdeep Singh Bajaj, Ali Elborady, Gianni del Bimbo, Mehrazin Alizadeh, David Montero, Pablo Martin-Ramiro, Muhammad Ibrahim, Oussama Tahiri Alaoui, John Malcolm, Samuel Mugel, Roman Orus  ( 2 min )

Changxiao Cai, T. Tony Cai, Hongzhe Li  ( 2 min )

Ashok Dahal, Rapha\"el Huser, Luigi Lombardo  ( 3 min )

Charles Meyers, Mohammad Reza Saleh Sedghpour, Tommy L\"ofstedt, Erik Elmroth  ( 3 min )

Levente Alekszejenk\'o, Tadeusz Dobrowiecki  ( 2 min )

Nikita Kozodoi, Elizaveta Zinovyeva, Simon Valentin, Jo\~ao Pereira, Rodrigo Agundez  ( 2 min )

Thomas Baldwin-McDonald, Mauricio A. \'Alvarez  ( 2 min )

Marios Papachristou, M. Amin Rahimian  ( 3 min )

Zhongjie Shi, Fanghui Liu, Yuan Cao, Johan A.K. Suykens  ( 3 min )

Zehua Liu, Zimeng Li, Jingyuan Wang, Yue He  ( 2 min )

Christian Yeo  ( 2 min )

Nicholas Galioto, Harsh Sharma, Boris Kramer, Alex Arkady Gorodetsky  ( 2 min )

Chengpiao Huang, Kaizheng Wang  ( 2 min )

Andrea Pugnana, Lorenzo Perini, Jesse Davis, Salvatore Ruggieri  ( 2 min )

Shoaib Meraj Sami, Md Mahedi Hasan, Nasser M. Nasrabadi, Raghuveer Rao  ( 3 min )

Vladan Pankovic  ( 3 min )

Vikram H. Zaveri  ( 2 min )

Vikram H. Zaveri  ( 3 min )

Valerie Poynor, Athanasios Kottas  ( 3 min )

Takao Yamaguchi  ( 2 min )

Mikhail Khovanov, Aaron D. Lauda, Joshua Sussan, Yasuyoshi Yonezawa  ( 2 min )

Tyler D. Robinson  ( 2 min )

Kirill Bykov, Laura Kopf, Shinichi Nakajima, Marius Kloft, Marina M.-C. H\"ohne  ( 2 min )

Paul Geuchen, Thomas Heindl, Dominik St\"oger, Felix Voigtlaender  ( 2 min )

Vincent Froese, Christoph Hertrich  ( 2 min )

Runtian Zhai, Bingbin Liu, Andrej Risteski, Zico Kolter, Pradeep Ravikumar  ( 3 min )

Seong Jin Cho, Gwangsu Kim, Junghyun Lee, Jinwoo Shin, Chang D. Yoo  ( 2 min )

Pierfrancesco La Mura, Yoav Shoham  ( 2 min )

Pierfrancesco La Mura  ( 2 min )

Pierfrancesco La Mura  ( 2 min )

Adam Brandenburger, Pierfrancesco La Mura  ( 2 min )

Tran Vu Khanh, Andrew Raich  ( 2 min )

N. Krivulin  ( 2 min )

Theodore Th. Voronov  ( 2 min )

Wojciech Tarnowski, Izaak Neri, Pierpaolo Vivo  ( 3 min )

Leonardo Enrique Abbrescia, Willie Wai Yeung Wong  ( 2 min )

Artur M. Schweidtmann, Dominik Bongartz, Daniel Grothe, Tim Kerkenhoff, Xiaopeng Lin, Jaromil Najman, Alexander Mitsos  ( 2 min )

Andr'e Unterberger  ( 2 min )

Thorsten Gl\"usenkamp  ( 3 min )

Jing An, Jianfeng Lu  ( 2 min )

Lorenzo Valleggi, Marco Scutari, Federico Mattia Stefanini  ( 3 min )

Naoki Egami, Musashi Hinck, Brandon M. Stewart, Hanying Wei  ( 3 min )

Leihao Chen, Onno Zoeter, Joris M. Mooij  ( 2 min )

Sascha Marton, Stefan L\"udtke, Christian Bartelt, Andrej Tschalzev, Heiner Stuckenschmidt  ( 2 min )

Heishiro Kanagawa, Alessandro Barp, Arthur Gretton, Lester Mackey  ( 2 min )

Emmanuel Caron, Stephane Chretien  ( 3 min )

Prateeti Mukherjee, Satya Lokam  ( 3 min )

Pedram Bakhtiarifard, Christian Igel, Raghavendra Selvan  ( 2 min )

S\"oren Schleibaum, J\"org P. M\"uller, Monika Sester  ( 3 min )

Philipp Schiele, Christoph Berninger, David R\"ugamer  ( 2 min )

Matthew Trager, Pramuditha Perera, Luca Zancato, Alessandro Achille, Parminder Bhatia, Stefano Soatto  ( 2 min )

Amir Akbarnejad, Gilbert Bigras, Nilanjan Ray  ( 3 min )

Zhikang Dong, Pawel Polak  ( 2 min )

Ziang Li, Yiwen Guo, Haodi Liu, Changshui Zhang  ( 2 min )

Manu Joseph, Harsh Raj  ( 2 min )

Morteza Haghir Chehreghani, Mostafa Haghir Chehreghani  ( 2 min )

Damir Filipovic, Michael Multerer, Paul Schneider  ( 2 min )

Ruichen Jiang, Aryan Mokhtari  ( 3 min )

Lorenzo Carlucci, Leszek Aleksander Ko{\l}odziejczyk, Francesco Lepore, Konrad Zdanowski  ( 2 min )

Paul Breiding, Peter B\"urgisser, Antonio Lerario, L\'eo Mathis  ( 2 min )

Tom Coates, Alexander Givental, Hsian-Hua Tseng  ( 2 min )

Hamidreza Kasaei, Songsong Xiong  ( 3 min )

Rafael B. Audibert, Henrique Lemos, Pedro Avelar, Anderson R. Tavares, Lu\'is C. Lamb  ( 3 min )

Wenhan Xia, Chengwei Qin, Elad Hazan  ( 2 min )

Yunfei Yang, Ding-Xuan Zhou  ( 2 min )

Dionissios T.Hristopulos  ( 3 min )

Louis Kirsch, James Harrison, Jascha Sohl-Dickstein, Luke Metz  ( 2 min )

Hyunouk Ko, Xiaoming Huo  ( 2 min )