Product
Help
Pricing
About
Jobs
Log in
Get started
Author
Add to collection
Liam Collins
Observed name:
Collins, Liam
Institution:
Snap (United States)
Observed institution:
Snap (United States)
ORCID:
0009-0006-3139-3339
Observed ORCID:
0009-0006-3139-3339
OpenAlex ID:
A5133136768
Implicit Reasoning for Large Language Model-based Generative Recommendation
2026
·
He, Yinhan
, Collins, Liam
, et al.
·
arXiv (Cornell University)
·
6
PDF
CoSearch: Joint Training of Reasoning and Document Ranking via Reinforcement Learning for Agentic Search
2026
·
Zeng, Hansi
, Collins, Liam
, et al.
·
arXiv (Cornell University)
·
5
PDF
Beyond Fixed Depths and Widths: Optimizing Textual Decoding Tries in LLM-based Generative Recommendation
2026
·
Jingzhe Liu
, Hanbing Wang
, et al.
·
ACM Conference on Recommender Systems (RecSys)
·
3
LLM-Based Generative Retrieval for Snapchat Content Recommendation
2026
·
Collins, Liam
, Ren, Jiwen
, et al.
·
arXiv (Cornell University)
·
2
PDF
Preserving Item Semantics for Free: Rethinking Token Initialization in LLM-Based Generative Recommendation
2026
·
Loveland, Donald
, Collins, Liam
, et al.
·
arXiv (Cornell University)
·
1
PDF
On the Memorization Behavior of LLMs in Generative Recommendation: Observations, Implications, and Training Strategies
2026
·
Kim, Sunwoo
, Lee, Sunkyung
, et al.
·
arXiv (Cornell University)
·
1
PDF
Understanding ID-Text Complementarity in Sequential Recommendation
2025
·
Collins, Liam
, Kumar, Bhuvesh
, et al.
·
arXiv (Cornell University)
·
1
PDF
Training-Free LLM-Based Recommendation with Post-LLM Item Refinement Using Collaborative Signals
2026
·
Kyungho Kim
, Sunwoo Kim
, et al.
·
ACM International Conference on Information and Knowledge Management (CIKM)
Hierarchical Token Prepending: Enhancing Information Flow in Decoder-based LLM Embeddings
2026
·
Association for Computational Linguistics 2026
, Akoglu, Leman
, et al.
·
Underline Science Inc.
Retrieval-Augmented Generation for LLM-based Recommender Systems: A Comprehensive Survey
2026
·
Kim, Sunwoo
, Lee, Geon
, et al.
·
HAL
PDF
Show more works
View as search
Works count:
10
Citation count:
19
H-index:
3
I10-index:
Year
Topic
Type