@vbecaert

Bridging research and real life, Sr Director Research ServiceNow Research

 Montréal québec
Joined November 2014
  • Tweets282
  • Following155
  • Followers1.1K
53Media
Valérie Bécaert retweeted
Torsten Scholak
@tscholak11 Apr 2025
🚨 SLAM Labs presents Apriel-5B! And it lands right in the green zone 🚨 Speed ⚡ + Accuracy 📈 + Efficiency 💸 This model punches above its weight, beating bigger LLMs while training on a fraction of the compute. Built with Fast-LLM, our in-house training stack. 🧵👇
5
48
8
133
14,919
Valérie Bécaert retweeted
Joan Rodriguez
@joanrod_ai20 Mar 2025
I’m excited to announce that 💫StarVector has been accepted at CVPR 2025! Over a year in the making, StarVector opens a new paradigm for Scalable Vector Graphics (SVG) generation by harnessing multimodal LLMs to generate SVG code that aesthetically mirrors input images and text. With this milestone, we’re also releasing StarVector on Hugging Face! 🥳🚀
12
56
12
175
16,422
We are hiring!!!! Managers and researchers. Many reasons to join us: ❤️The people are just fantastic 🚀We do open source (The Stack, Starcoder, FastLLM, TapeAgents, BigDocs, ...) 🙀We reach millions of users
ServiceNow AI Research
@ServiceNowRSRCH9 Dec 2024
Replying to @ServiceNowRSRCH
Manager, Research Scientist Management Apply today: careers.servicenow.com/jobs/… As a manager, you will manage, coach and guide our researchers, specifically focusing on generative AI. You will support the research lead of a research Lab or program in building the team’s roadmap and achieving the team's goal. Additionally, you may conduct your research or participate in ongoing research projects. Your contributions to knowledge generation and transfer will help maintain our scientific credibility within the academic community and our impact on the ServiceNow products. What you get to do in this role: You will be responsible for managing a team of 5 to 10 researchers and engineers, in collaboration with the research leads.  You will be responsible of the growth, development  and wellbeing of the team and will provide tactical and operational support to the research leads. You will have the opportunity to conduct your own research or participate to transfer activities of the AI Research team.
4
8
934
Manager, Research Scientist Management Apply today: careers.servicenow.com/jobs/… As a manager, you will manage, coach and guide our researchers, specifically focusing on generative AI. You will support the research lead of a research Lab or program in building the team’s roadmap and achieving the team's goal. Additionally, you may conduct your research or participate in ongoing research projects. Your contributions to knowledge generation and transfer will help maintain our scientific credibility within the academic community and our impact on the ServiceNow products. What you get to do in this role: You will be responsible for managing a team of 5 to 10 researchers and engineers, in collaboration with the research leads.  You will be responsible of the growth, development  and wellbeing of the team and will provide tactical and operational support to the research leads. You will have the opportunity to conduct your own research or participate to transfer activities of the AI Research team.
1
1
3
3
3,898
Congratulations to the #ICML2024 best paper award winners. Special shout out to @DjDvij on contributions to 'Stealing Part of a Production Language Model'. While DJ did not complete this work at ServiceNow, he recently joined our team as the research lead for Safety & Security.
ICML Conference@icmlconf23 Jul 2024
Congratulations to the best paper award winners
2
7
1,040
Valérie Bécaert retweeted
Catherine Martin@cathe_martin11 Dec 2023
Super happy to be at #NeurIPS23 again! Come see our team @ServiceNowRSRCH booth.
1
10
32
2,007
Congratulations to all the @BigCodeProject contributors that worked tirelessly over the last 6+ months to bring the vision of releasing a responsibly developed 15B parameter Code LLM to fruition. We cannot thank you enough for the collaboration & contributions to the community.
BigCode@BigCodeProject4 May 2023
Replying to @BigCodeProject
We present the most extensive evaluation of code LLMs to date in the full tech report with 68 (!) authors. You can also read up on all the details from data preprocessing and governance to training at scale! drive.google.com/file/d/1cN-…
6
31
2,280
Valérie Bécaert retweeted
Nicolas Chapados@NicolasChapados5 May 2023
Changing the way Generative AI can help developers write better code quicker: introducing StarCoder, a top-performing responsibly-built open-science 15B-parameter LLM for code. Amazing collaboration between @ServiceNowRSRCH , @huggingface & a TEAM of 100s of global contributors!
BigCode@BigCodeProject4 May 2023
Introducing: 💫StarCoder StarCoder is a 15B LLM for code with 8k context and trained only on permissive data in 80+ programming languages. It can be prompted to reach 40% pass@1 on HumanEval and act as a Tech Assistant. Try it here: shorturl.at/cYZ06r Release thread🧵
2
4
1
25
2,954
100% agree. You learned to activate your creativity and treat your ideas with scientific rigour
Michael Black
@Michael_J_Black30 Mar 2023
PhD students, don't worry. Technologies, trends, and even whole fields come and go. A PhD makes you an expert in a field but, more importantly, teaches you how to become an expert. Once you know that you can learn anything, you can adapt to major disruptions in your field.
1
471
Valérie Bécaert retweeted
Travis Darrow@TravisDarrow22 Nov 2022
Join ServiceNow's Valérie Bécaert and Dorit Levy-Zilbershot at the Women in AI Flagship Summit at the AI Summit NYC on 12/8 where they’ll discuss building diverse teams and a collaborative culture to advance AI. via @FalkMarques @business_ai #AISummit @women_in_ai @vbecaert
2
Valérie Bécaert retweeted
BigCode@BigCodeProject1 Dec 2022
Today we are releasing The Stack v1.1! 🚀 We added more data, included more programming languages, and extended the list of permissive licenses used. huggingface.co/datasets/bigc… Also the first batch of opt-out requests was removed from the dataset.
2
42
6
159
Valérie Bécaert retweeted
VentureBeat
@VentureBeat4 Nov 2022
As code #LLMs mature, we’re beginning to discover the challenges they must overcome, including licensing, transparency, security and control. See how #BigCode, @HuggingFace and @ServiceNow are collaborating on these issues: venturebeat.com/ai/how-code-…
5
1
10
🪅 Congratulations to the @BigCodeProject research community on releasing 📑The Stack, a 3TB #dataset of permissively licensed code in 30 programming languages for pretraining large language models for code. #OpenScience #AI #Research #LLM #Code Read more servicenow.com/blogs/2022/bi…
The Stack - 3 TB of permissive code data - With love from @BigCodeProject

ALT The Stack - 3 TB of permissive code data - With love from @BigCodeProject

The Stack dataset collection process is outlined in this diagram. Dataset collection took several months to complete. To create The Stack, 220.92 million unique GitHub source code repository names were collected from GH Archive, with 51.76 billion files (and 5.28 billion unique files) successfully downloaded from 137.36 million public and accessible repositories. The uncompressed size of all stored files is 92.36TB.

ALT The Stack dataset collection process is outlined in this diagram. Dataset collection took several months to complete. To create The Stack, 220.92 million unique GitHub source code repository names were collected from GH Archive, with 51.76 billion files (and 5.28 billion unique files) successfully downloaded from 137.36 million public and accessible repositories. The uncompressed size of all stored files is 92.36TB.

This image shows a histogram of the amount of data per programming language for the permissive license dataset. Note that we plot the dataset on a log scale. For the permissive-license dataset, the four biggest code bases are HTML (746GB), JavaScript (486GB), Java (271GB), and C (22GB). These comprise more than 55% of the total dataset size.

ALT This image shows a histogram of the amount of data per programming language for the permissive license dataset. Note that we plot the dataset on a log scale. For the permissive-license dataset, the four biggest code bases are HTML (746GB), JavaScript (486GB), Java (271GB), and C (22GB). These comprise more than 55% of the total dataset size.

This chart shows The Stack dataset compared to CodeParrot, AlphaCode, CodeGen, and PolyCoder. The Stack and CodeParrot provide source code for 30 programming languages, while AlphaCode, PolyCoder, and CodeGen provide source code data for only 12, 12, and 6 languages, respectively. The Stack is more than three times the size of CodeParrot, the next-largest publicly available code dataset, and is also bigger than CodeParrot for each individual programming language.

ALT This chart shows The Stack dataset compared to CodeParrot, AlphaCode, CodeGen, and PolyCoder. The Stack and CodeParrot provide source code for 30 programming languages, while AlphaCode, PolyCoder, and CodeGen provide source code data for only 12, 12, and 6 languages, respectively. The Stack is more than three times the size of CodeParrot, the next-largest publicly available code dataset, and is also bigger than CodeParrot for each individual programming language.

21
1
31
Valérie Bécaert retweeted
Amy Daali, PhD@wdaali99916 Aug 2022
Excited to announce our #SXSW 2023 Panel Picker "The Digital Twin Inside Us All"! Together with @devonbleibtrey and @vbecaert, we will discuss how #DigitalTwins will be transforming our lives. VOTE here 👉: panelpicker.sxsw.com/vote/12…
1
2
Valérie Bécaert retweeted
Devon Bleibtrey
@devonbleibtrey10 Aug 2022
Together with @wdaali999 and @vbecaert, we submitted a panel about #DigitalTwins and the way they will be transforming our lives. To help our panel ‘The Digital Twin Inside Us All’ present at SXSW, you can cast your vote here: panelpicker.sxsw.com/vote/12…
6
3
15
Valérie Bécaert retweeted
Alexandre Lacoste@alex_lacoste_19 Jun 2022
The most important is that there are things we can do @wellingmax. If you want to know how our ML community can help, please follow @ClimateChangeAI . Specifically, you can read "Tackling Climate Change with Machine Learning" :
Max Welling@wellingmax18 Jun 2022
16 experts were interviewed on global warning. All said that it will be above 2C. 60% thought between 3 and 5. At 4 degrees large parts of the planet are inhabitable by the end of the century. Your children and grandchildren will experience the resulting chaos. Think about that.
1
1
7
Valérie Bécaert retweeted
Hector Palacios@hectorpal20 Jun 2022
We won! What? ALFRED is agent in 3D environment receiving tasks in natural langue. How? Symbolic planning + DL perception. Why? The fundamental problem is a re-taskable agent with fixed skills and ontology. Who? Xiaotian Lui, intern at @ServiceNowRSRCH , co-adviced with @cjmuise
MuLab@QuMuLab20 Jun 2022
Very happy to share the news that recent @QueensComputing graduate, Xiaotian Liu, co-advised by @hectorpal (ServiceNow) & @cjmuise (Queen's) has won the #CVPR2022 #EAI22 embodied agent ALFRED challenge: askforalfred.com/EAI22/

ALFRED

A Benchmark for Interpreting Grounded Instructions for Everyday Tasks

askforalfred.com
2
13
Popular image datasets assign a single label to samples that contain multiple objects. A solution is to relabel them. But what if the dataset is too big to relabel? Multi-label Iterated Learning (MILe) is inspired by the evolution of language. blogs.servicenow.com/2022/mu… [1/5]
1
14
19
About connecting research to product. Recording of the NVIDIA GTC - Leaders in AI Panel. @NicolasChapados has interesting things to say!!!
ServiceNow AI Research
@ServiceNowRSRCH23 Mar 2022
Replying to @ServiceNowRSRCH @nvidia @NicolasChapados @ctnzr @jure @NVIDIAGTC
ICYMI Here's the link to the recording and highlights from the Q&A for the NVIDIA GTC - Leaders in AI Panel Discussion that took place Tuesday, March 22, 2022. linkedin.com/pulse/nvidia-gt…
Leaders in AI Panel Discussion: AI Journey from Concept to Scale NVIDIA GTC 2022, Tuesday, March 22, 12:00 PM - 12:50 PM PDT Nicolas Chapados (VP Research at ServiceNow), met online with Bryan Catanzaro (VP Applied Deep Learning Research at NVIDIA), Ya Xu (VP of Engineering and Head of Data at LinkedIn), and Jure Leskovec (Associate Professor and Senior Technical Advisor at Stanford University and Pinterest). Bryan Catanzaro led the discussion with the panel and explored how each of the leaders is bridging the gap between AI research and production, their approaches to technology transfer between research and engineering, recent achievements in AI, and the future of AI. Throughout the GTC session broadcast, there was a real-time Q&A, with Bryan Catanzaro and Nicolas Chapados both moderating. You'll need to register and watch the session replay online to get the full story, but we’ve highlighted a few of the questions that Nicolas answered during the Q&A for your reading pleasure.

ALT Leaders in AI Panel Discussion: AI Journey from Concept to Scale NVIDIA GTC 2022, Tuesday, March 22, 12:00 PM - 12:50 PM PDT Nicolas Chapados (VP Research at ServiceNow), met online with Bryan Catanzaro (VP Applied Deep Learning Research at NVIDIA), Ya Xu (VP of Engineering and Head of Data at LinkedIn), and Jure Leskovec (Associate Professor and Senior Technical Advisor at Stanford University and Pinterest). Bryan Catanzaro led the discussion with the panel and explored how each of the leaders is bridging the gap between AI research and production, their approaches to technology transfer between research and engineering, recent achievements in AI, and the future of AI. Throughout the GTC session broadcast, there was a real-time Q&A, with Bryan Catanzaro and Nicolas Chapados both moderating. You'll need to register and watch the session replay online to get the full story, but we’ve highlighted a few of the questions that Nicolas answered during the Q&A for your reading pleasure.

3
Hey! See that? New name same great research !!
ServiceNow AI Research
@ServiceNowRSRCH28 Jan 2022
1/10 You may have noticed a few changes on our channel today! It’s been a year since the acquisition of Element AI by @ServiceNow. While we have given our account a new name, we’re still as committed as ever to making socially responsible contributions to the AI community.
1/10 You may have noticed a few changes on our channel today! It’s been a year since the acquisition of Element AI by ServiceNow. While we have given our account a new name, we’re still as committed as ever to making socially responsible contributions to the AI community. #ServiceNow #ElementAI #ServiceNowNews #News #AI #ArtificialIntelligence #ML #MachineLearning #ComputerVision #Research #AIResearch

ALT 1/10 You may have noticed a few changes on our channel today! It’s been a year since the acquisition of Element AI by ServiceNow. While we have given our account a new name, we’re still as committed as ever to making socially responsible contributions to the AI community. #ServiceNow #ElementAI #ServiceNowNews #News #AI #ArtificialIntelligence #ML #MachineLearning #ComputerVision #Research #AIResearch

5
Load more