Groups
Conversations
All groups and messages
Send feedback to Google
Help
Training
Sign in
Groups
Machine Learning Reading Group Yerevan
Conversations
About
Machine Learning Reading Group Yerevan
Contact owners and managers
1–30 of 230
Mark all as read
Report abusive group
0 selected
Tatev Vardanyan
Dec 6
Meeting #178 | Language Is Not All You Need: Aligning Perception with Language Models
Dear Friends, On Friday, December 8th, at 3:30 pm, at the YSU-Krisp AI Lab, YerevaNN's ML
unread,
Meeting #178 | Language Is Not All You Need: Aligning Perception with Language Models
Dear Friends, On Friday, December 8th, at 3:30 pm, at the YSU-Krisp AI Lab, YerevaNN's ML
Dec 6
Tatev Vardanyan
Nov 30
Meeting #177 | On Bringing Robots Home
Dear community, Tomorrow, December 1, at 3:30 pm, at the YSU-Krisp AI Lab, YerevaNN's ML
unread,
Meeting #177 | On Bringing Robots Home
Dear community, Tomorrow, December 1, at 3:30 pm, at the YSU-Krisp AI Lab, YerevaNN's ML
Nov 30
Tatev Vardanyan
Nov 22
Meeting #176 | Cooperative Deep-Learning Positioning in mmWave 5G-Advanced Networks
Good morning Friends, Tomorrow, November 24, at 3:30 pm, at the YSU-Krisp AI Lab, YerevaNN's ML
unread,
Meeting #176 | Cooperative Deep-Learning Positioning in mmWave 5G-Advanced Networks
Good morning Friends, Tomorrow, November 24, at 3:30 pm, at the YSU-Krisp AI Lab, YerevaNN's ML
Nov 22
Tatev Vardanyan
Nov 16
Meeting #175 | What Can Transformers Learn In-Context? A Case Study of Simple Function Classes
Dear friends, On Friday, NOVEMBER 17, at 3:30 pm, YerevaNN's ML Researcher Qnarik Mheryan will
unread,
Meeting #175 | What Can Transformers Learn In-Context? A Case Study of Simple Function Classes
Dear friends, On Friday, NOVEMBER 17, at 3:30 pm, YerevaNN's ML Researcher Qnarik Mheryan will
Nov 16
Tatev Vardanyan
Nov 9
Meeting #174 | Vision Transformers Need Registers + Efficient Streaming Language Models with Attention Sinks
Dear friends, On Friday, NOVEMBER 10, at 3:30 pm, YerevaNN's ML Researcher Philipp Guevorguian
unread,
Meeting #174 | Vision Transformers Need Registers + Efficient Streaming Language Models with Attention Sinks
Dear friends, On Friday, NOVEMBER 10, at 3:30 pm, YerevaNN's ML Researcher Philipp Guevorguian
Nov 9
Tatev Vardanyan
Nov 1
Meeting #173 | Towards Last-layer Retraining for Group Robustness with Fewer Annotations
Dear Friends, On Friday, NOVEMBER 3rd, at 3:30 pm, Rafayel Darbinyan, YerevaNN ML Researcher, will
unread,
Meeting #173 | Towards Last-layer Retraining for Group Robustness with Fewer Annotations
Dear Friends, On Friday, NOVEMBER 3rd, at 3:30 pm, Rafayel Darbinyan, YerevaNN ML Researcher, will
Nov 1
Tatev Vardanyan
Oct 24
Meeting #172 | ICCV23 Insights and Takeaways from YerevaNN Researchers
Dear friends, This month, YerevaNN researchers Vahan Huroyan, Ani Vanyan, and Alla Barseghyan
unread,
Meeting #172 | ICCV23 Insights and Takeaways from YerevaNN Researchers
Dear friends, This month, YerevaNN researchers Vahan Huroyan, Ani Vanyan, and Alla Barseghyan
Oct 24
Tatev Vardanyan
Oct 20
Meeting #171 | DClEVerNet: Deep Combinatorial Learning for Efficient EV Charging Scheduling in Large-scale Networked Facilities
Dear friends, Join us at 3:30 today, October 20, at YSU-Krisp AI Lab for a seminar by Khazhak
unread,
Meeting #171 | DClEVerNet: Deep Combinatorial Learning for Efficient EV Charging Scheduling in Large-scale Networked Facilities
Dear friends, Join us at 3:30 today, October 20, at YSU-Krisp AI Lab for a seminar by Khazhak
Oct 20
Hrant Khachatrian
Oct 6
Lecture series "Interpretable models: the challenge ahead" by Nelson Baloian
Dear friends, Today and the two subsequent Fridays Nelson Baloian from University of Chile will
unread,
Lecture series "Interpretable models: the challenge ahead" by Nelson Baloian
Dear friends, Today and the two subsequent Fridays Nelson Baloian from University of Chile will
Oct 6
Tatev Vardanyan
Sep 27
Meeting #170 | PyTorch FSDP: Experiences on Scaling Fully Sharded Data Parallel
Dear Friends, On Friday, September 29, Menua Bedrossian of YerevaNN will present a paper titled
unread,
Meeting #170 | PyTorch FSDP: Experiences on Scaling Fully Sharded Data Parallel
Dear Friends, On Friday, September 29, Menua Bedrossian of YerevaNN will present a paper titled
Sep 27
Tatevik Vardanyan
Sep 21
Meeting #169 | DOLA: Decoding by Contrasting Layers Improves Factuality in LLMs
Dear Friends, Tomorrow's seminar shall be delivered by Ani Tevosyan, YerevaNN researcher. Ani
unread,
Meeting #169 | DOLA: Decoding by Contrasting Layers Improves Factuality in LLMs
Dear Friends, Tomorrow's seminar shall be delivered by Ani Tevosyan, YerevaNN researcher. Ani
Sep 21
Tatevik Vardanyan
2
Sep 14
Meeting #168 | Statistical and Computational Complexity of the Feature Matching Map Detection Problem
Dear Friends, This is a kind reminder about tomorrow's seminar. Please find details in the
unread,
Meeting #168 | Statistical and Computational Complexity of the Feature Matching Map Detection Problem
Dear Friends, This is a kind reminder about tomorrow's seminar. Please find details in the
Sep 14
Hrant Khachatrian
Sep 6
DataFest Yerevan 2023
Dear colleagues, There will be no reading group meeting this week. Instead we invite you all to
unread,
DataFest Yerevan 2023
Dear colleagues, There will be no reading group meeting this week. Instead we invite you all to
Sep 6
Tatevik Vardanyan
Aug 31
Meeting #167 | Regression Transformer enables concurrent sequence regression and generation for molecular language modelling
Dear Community, Tomorrow, September 1, from 3:30 pm to 5:00 pm, YerevaNN ML researcher Gayane
unread,
Meeting #167 | Regression Transformer enables concurrent sequence regression and generation for molecular language modelling
Dear Community, Tomorrow, September 1, from 3:30 pm to 5:00 pm, YerevaNN ML researcher Gayane
Aug 31
Tatevik Vardanyan
Aug 24
Meeting #166: Spatial Transform Decoupling for Oriented Object Detection
Dear friends, Tomorrow, on August 25, YerevanNN ML researcher Ani Vanyan will present a paper on
unread,
Meeting #166: Spatial Transform Decoupling for Oriented Object Detection
Dear friends, Tomorrow, on August 25, YerevanNN ML researcher Ani Vanyan will present a paper on
Aug 24
Tatevik Vardanyan
,
Hrant Khachatrian
3
Aug 15
Meeting #164: ICML 2023: Notes and Insights
Dear Friends, Hrant Khachatrian will continue the seminar by sharing his takeaways and notes from
unread,
Meeting #164: ICML 2023: Notes and Insights
Dear Friends, Hrant Khachatrian will continue the seminar by sharing his takeaways and notes from
Aug 15
Tatevik Vardanyan
Aug 1
Meeting #163: High-performance brain-to-text communication via handwriting
Dear Community, On Friday, at 3:30, Alla Barseghyan of YerevaNN will present a paper on brain-
unread,
Meeting #163: High-performance brain-to-text communication via handwriting
Dear Community, On Friday, at 3:30, Alla Barseghyan of YerevaNN will present a paper on brain-
Aug 1
Machine Learning Reading Group Yerevan
Jul 25
Meeting #162: [YSU AI Lab; Friday 16:00] Language Models Don't Always Say What They Think
Hi everyone, This week we'll delve into the captivating research titled "Language Models Don
unread,
Meeting #162: [YSU AI Lab; Friday 16:00] Language Models Don't Always Say What They Think
Hi everyone, This week we'll delve into the captivating research titled "Language Models Don
Jul 25
Tatevik Vardanyan
Jul 19
Meeting #161: CoTracker -- It is Better to Track Together
On Friday, July 21, at 3:30 pm, Hakob Tamazyan of YerevaNN shall present a paper titled CoTracker: It
unread,
Meeting #161: CoTracker -- It is Better to Track Together
On Friday, July 21, at 3:30 pm, Hakob Tamazyan of YerevaNN shall present a paper titled CoTracker: It
Jul 19
Hrant Khachatrian
Jul 7
Meeting #160: VQ-VAE | Neural Discrete Representation Learning
Hi everyone, Today at 15.30 Vahan Huroyan will present a 2017 paper on Neural Discrete Representation
unread,
Meeting #160: VQ-VAE | Neural Discrete Representation Learning
Hi everyone, Today at 15.30 Vahan Huroyan will present a 2017 paper on Neural Discrete Representation
Jul 7
Tatevik Vardanyan
2
Jun 29
Meeting #159: On message broadcasting in interconnection networks
Ընկերներ վաղվա սեմինարը տեղափոխվել է ժամը 14:00-ին: Friends, tomorrow's seminar has been
unread,
Meeting #159: On message broadcasting in interconnection networks
Ընկերներ վաղվա սեմինարը տեղափոխվել է ժամը 14:00-ին: Friends, tomorrow's seminar has been
Jun 29
Tatevik Vardanyan
2
Jun 22
AI Conference at YSU
Attached please find the agenda for the AI Conf 2023. Regards, Tatevik On Thu, Jun 22, 2023, 17:46
unread,
AI Conference at YSU
Attached please find the agenda for the AI Conf 2023. Regards, Tatevik On Thu, Jun 22, 2023, 17:46
Jun 22
Hrant Khachatrian
Apr 20
No meeting this week. + AINL and DS meetup
Hi everyone, Today AINL conference started at RAU. It will be held for two more days. At 10am on
unread,
No meeting this week. + AINL and DS meetup
Hi everyone, Today AINL conference started at RAU. It will be held for two more days. At 10am on
Apr 20
Hrant Khachatrian
Apr 6
Meeting #153: [YSU AI Lab; Friday 15:30] Unifying Molecular and Textual Representations via Multi-task Language Modelling
Hello everyone, This Friday at 3:30 pm, Philipp Guevorguian of YerevaNN will present a paper Unifying
unread,
Meeting #153: [YSU AI Lab; Friday 15:30] Unifying Molecular and Textual Representations via Multi-task Language Modelling
Hello everyone, This Friday at 3:30 pm, Philipp Guevorguian of YerevaNN will present a paper Unifying
Apr 6
hr...@yerevann.com
Apr 6
Updated invitation: Meeting #153: Unifying Molecular and Textual Representati... @ Fri Apr 7, 2023 3:30pm - 5:30pm (AMT) (Machine Learning Reading Group Yerevan)
Meeting #153: Unifying Molecular and Textual Representations via Multi-task Language Modelling This
unread,
Updated invitation: Meeting #153: Unifying Molecular and Textual Representati... @ Fri Apr 7, 2023 3:30pm - 5:30pm (AMT) (Machine Learning Reading Group Yerevan)
Meeting #153: Unifying Molecular and Textual Representations via Multi-task Language Modelling This
Apr 6
Hrant Khachatrian
Mar 16
Meeting #150: [YSU AI Lab; Friday 15:30] On Improving Adversarial Transferability of Vision Transformers
Hello everyone, This Friday at 3:30 pm, Davit Papikyan from YerevaNN will present the paper "On
unread,
Meeting #150: [YSU AI Lab; Friday 15:30] On Improving Adversarial Transferability of Vision Transformers
Hello everyone, This Friday at 3:30 pm, Davit Papikyan from YerevaNN will present the paper "On
Mar 16
Karen Hambardzumyan
Mar 9
Meeting #149: [YSU AI Lab; Friday 15:30] GradSkip: Communication-Accelerated Local Gradient Methods with Better Computational Complexity
This Friday at 3:33 pm, Arto Maranjyan from YerevaNN will present his joint work with Mher Safaryan
unread,
Meeting #149: [YSU AI Lab; Friday 15:30] GradSkip: Communication-Accelerated Local Gradient Methods with Better Computational Complexity
This Friday at 3:33 pm, Arto Maranjyan from YerevaNN will present his joint work with Mher Safaryan
Mar 9
Karen Hambardzumyan
Mar 2
Meeting #148: [YSU AI Lab; Friday 15:30] DINO; Emerging Properties in Self-Supervised Vision Transformers
This Friday at 3:30 pm, Ani Vanyan from YerevaNN will present ICCV 2021 paper "Emerging
unread,
Meeting #148: [YSU AI Lab; Friday 15:30] DINO; Emerging Properties in Self-Supervised Vision Transformers
This Friday at 3:30 pm, Ani Vanyan from YerevaNN will present ICCV 2021 paper "Emerging
Mar 2
Hrant Khachatrian
Feb 24
Meeting #147: [YSU AI Lab; Friday 14.00] Toolformer: Language Models Can Teach Themselves to Use Tools
This week Ani Tevosyan will present a recent paper called ToolFormer. It's about language models
unread,
Meeting #147: [YSU AI Lab; Friday 14.00] Toolformer: Language Models Can Teach Themselves to Use Tools
This week Ani Tevosyan will present a recent paper called ToolFormer. It's about language models
Feb 24
Hrant Khachatrian
Feb 3
Meeting #144: [Offline; YSU AI Lab; Friday 16.00] A new method for domain generalization
Dear friends, Today Karen Hambardzumyan will present a recent paper about bridging pretrained textual
unread,
Meeting #144: [Offline; YSU AI Lab; Friday 16.00] A new method for domain generalization
Dear friends, Today Karen Hambardzumyan will present a recent paper about bridging pretrained textual
Feb 3