Delivering a report bringing you to the lively scene at NeurIPS 2022 X LG 
This yearās NeurIPS took place in New Orleans, LA in the US, the same place as CVPR!
The scene may look familiar for those of you who read our CVPR post. Five subsidiaries including LG AI Research participated in the NeurIPS conference (Woo!).
Many participants, including first authors of successful papers, were active on the scene at the conference. Letās head on over to the lively scene!
(A quick photo with everyone before opening the booth on the first day of the conference )
)
So many people visited our booth during the conference! The Global AI Center in Ann Arbor has gained tremendous interest, and our researchers in Korea are receiving a constant stream of inquiries for full-time employment and internships. (Itās no secret that the overwhelming number of applications crashed our recruitment website during the conference )
)
Hold on, letās take a moment to check out the LG AI Researcher Paper List! In this yearās NeurIPS, LG AI Research published seven papers and was actively involved in sessions with workshop papers as well!
Conference
1. CEDe: A collection of expert-curated datasets with atom-level entity annotations for Optical Chemical Structure Recognition
2. UniCLIP: Unified Framework for Contrastive LanguageāImage Pre-training
3. Transformers meet Stochastic Block Models: Attention with Data-Adaptive Sparsity and Cost
4. Unsupervised Visual Representation Learning via Mutual Information Regularized Assignment
5. Pure Transformers are Powerful Graph Learners
6. Transferring Pre-trained Multimodal Representations with Cross-modal Similarity Matching
7. LobsDICE: Offline Learning from Observation via Stationary Distribution Correction Estimation
Workshop
1. Dynamics-Augmented Decision Transformer for Offline Dynamics Generalization
2. ReSPack: A Large-Scale Rectilinear Steiner Tree Packing Data Generator and Benchmark
3. Allele-conditional attention mechanism for HLA-peptide complex binding affinity prediction
4. Learning Exploration Policies with View-based Intrinsic Rewards
5. Instrument Separation of Symbolic Music by Explicitly Guided Diffusion Model
Conference
1. CEDe: A collection of expert-curated datasets with atom-level entity annotations for Optical Chemical Structure Recognition
2. UniCLIP: Unified Framework for Contrastive LanguageāImage Pre-training
3. Transformers meet Stochastic Block Models: Attention with Data-Adaptive Sparsity and Cost
4. Unsupervised Visual Representation Learning via Mutual Information Regularized Assignment
5. Pure Transformers are Powerful Graph Learners
6. Transferring Pre-trained Multimodal Representations with Cross-modal Similarity Matching
7. LobsDICE: Offline Learning from Observation via Stationary Distribution Correction Estimation
Workshop
1. Dynamics-Augmented Decision Transformer for Offline Dynamics Generalization
2. ReSPack: A Large-Scale Rectilinear Steiner Tree Packing Data Generator and Benchmark
3. Allele-conditional attention mechanism for HLA-peptide complex binding affinity prediction
4. Learning Exploration Policies with View-based Intrinsic Rewards
5. Instrument Separation of Symbolic Music by Explicitly Guided Diffusion Model
As the papers impressive results, many researchers were able to join on this business trip. Our researchers promoted LG AI Research at our booth and gave outstanding presentations during the poster session. They truly are the best of the best!
(Our proud LG AI researcher giving a poster presentation!)
(So cool!)
On the afternoon of NovemberĀ 28, Honglak Lee, CSAI gave an Expo Talk on the topic of āTowards learning agents for solving complex real-world tasksā. Many people filled the large lecture room to attend Honglakās lecture! (As expected! )
)
The second day of the conference was the highly anticipated event that everyone had been waiting for!
LG AI DAY! (Yay!)
LG AI Day is a networking dinner event hosted by LG for first authors at top-tier global conferences! LG AI Day is a delightful event where we introduce AI research and the LG vision and engage in endless conversation while enjoying delicious a course meal and an open bar.
LG AI Day is a great place to network with not only LG employees but also other participants. This is probably why there was endless conversation on research during the meal, break time, and even during dessert.
(Everyone is focused on the raffle! Unfortunately, LG employees didnāt have a chance )
)We also had fun answering interesting quiz questions to finish the enjoyable time on an even happier note! There was also raffle for a chance to win incredible prizes!
Prizes included an LG StanbyME and an iPad Mini with LG Innotekās camera module Everyone was excited in anticipation, and the bright smiles of the winners are still vivid in our memories
Everyone was excited in anticipation, and the bright smiles of the winners are still vivid in our memories Everyone who contributed to the preparation of the event was proud to hear positive stories about LG AI Day throughout the conference.
Weāll be back with an even better event next time. See you again next time! Bye!