Collaborative Filtering Methods for Paper Recommendation Systems
Describing collaborative filtering methods from classical SVD to NeuMF/GraphNeuMF/DMF and an ensemble to recommend papers.
Posts by Collection
portfolio
The MCMC Inference Engine Behind a PPL
Overview of building an MCMC engine with practical implementation details.
KV-Cache Refresh Methods for Long Generation
In this blog, we show how helpful KV cache refreshes can be for long generation from small models, along with efficient ways of finding when to refresh using inference algorithms.
publications
Percutaneous cannulated screw fixation in the treatment for diabetic ankle fractures
Published in European Journal of Orthopedics, 2020
Comprehensive analysis of correlation between diabetic patients and ankle fractures
Inferring Interpretable Semantic Cognitive Maps from Noisy Document Corpora
Published in ICAART, 2024
Extracting large-scale semantic spaces from text corpora
Recommended citation: Emara et al. (2024). "Inferring Interpretable Semantic Cognitive Maps from Noisy Document Corpora." ICAART. 1(1).
Download Paper
Syntactic and Semantic Control of Large Language Models via Sequential Monte Carlo
Published in ICLR, 2025
Controlled LM generation based on sequential Monte Carlo
Recommended citation: Loula et al. (2025). "Syntactic and Semantic Control of Large Language Models via Sequential Monte Carlo." ICLR. 1(1).
Download Paper
Language Models over Canonical Byte-Pair Encodings
Published in ICML, 2025
Fixing LM probability leakage from unintuitive tokenizations, by adjusting LM distribution
Recommended citation: Vieira et al. (2025). "Language Models over Canonical Byte-Pair Encodings." ICML. 1(1).
Download Paper
talks
Language Models and Using Instagram Data for Investment Opportunities
Published:
Language Models are now heavily used in Investment, but a big portion of data analysis done by companies in the financial sector has not fully integrated its capabilites. In this presentation, we show different methods of obtaining customer experience data on Instagram and present a solution to link it to the performance of a company’s stock, providing information on whether to invest or not
teaching
Natural Language Processing
Masters course, ETH Zurich, 2024
Teaching Assistant for Natural Language Processing Class; Taught by Prof. Ryan Cotterell