Gustavo Penha

650 posts

Gustavo Penha banner
Gustavo Penha

Gustavo Penha

@_Guz_

Research Scientist @Spotify · Working with IR, RecSys, NLP · PhD from @tudelft · ex @AmazonScience · https://t.co/SMu8BlyfIb

Holanda (Países Baixos) Katılım Ocak 2009
564 Takip Edilen823 Takipçiler
Gustavo Penha retweetledi
Aixin Sun
Aixin Sun@AixinSG·
I doubt to what extent improvements on these datasets would translate to improvements in today's real-world recommendation settings. Reference: arxiv.org/abs/2508.19399…
Aixin Sun tweet media
English
1
4
10
916
Gustavo Penha
Gustavo Penha@_Guz_·
@svakulenk0 Thanks! But I won’t go this time :( my co-authors will present this time
English
0
0
0
19
Gustavo Penha
Gustavo Penha@_Guz_·
Happy to share our #recsys25 paper: “Evaluating Podcast Recommendations with Profile-Aware LLM-as-a-Judge”. 🧠 90 days of listening → natural-language user profiles → LLM judges alignment 📊 Aligns with human eval. With amazing Spotify co-authors. 📄 arxiv.org/abs/2508.08777
English
2
5
13
1K
Gustavo Penha
Gustavo Penha@_Guz_·
@svakulenk0 Yes! We calculate the agreement between the llm judge and users in the paper
English
2
0
1
42
Gustavo Penha
Gustavo Penha@_Guz_·
Excited to share our paper “Semantic IDs for Joint Generative Search & Recommendation” @ RecSys'25 🧠 Jointly fine-tuning embeddings for both tasks → shared Semantic IDs that work for search and recs ⚖️ 📦 No more task-specific trade-offs!
Gustavo Penha tweet media
English
2
3
8
485
Gustavo Penha retweetledi
Sumit
Sumit@_reachsumit·
Semantic IDs for Joint Generative Search and Recommendation @_Guz_ et al. at Spotify introduce a bi-encoder model fine-tuned on both search and recommendation tasks to obtain item embeddings, followed by construction of unified Semantic ID space. 📝arxiv.org/abs/2508.10478
English
0
2
11
687
Gustavo Penha retweetledi
Sumit
Sumit@_reachsumit·
Evaluating Podcast Recommendations with Profile-Aware LLM-as-a-Judge Spotify introduces a profile-aware LLM framework for evaluating personalized podcast recommendations using natural-language user profiles distilled from listening history. 📝arxiv.org/abs/2508.08777
English
0
6
22
1.2K
Gustavo Penha retweetledi
Sumit
Sumit@_reachsumit·
Describe What You See with Multimodal Large Language Models to Enhance Video Recommendations @denadai2 et al. at Spotify use multimodal LLMs to generate natural-language descriptions of video content for better recommendations 📝arxiv.org/abs/2508.09789 👨🏽‍💻huggingface.co/datasets/marco…
English
0
4
8
610
Gustavo Penha retweetledi
Marco De Nadai
Marco De Nadai@denadai2·
What if we could use off-the-shelf Multimodal Large Language Model to enrich current video recommendation models? This is what we asked ourselves in our recent #recsys2025 paper arxiv.org/pdf/2508.09789 🧵
Marco De Nadai tweet media
English
1
2
6
420
Gustavo Penha
Gustavo Penha@_Guz_·
🔎 LLM alignment techniques can enhance query expansion by eliminating the need for multiple generations followed by re-ranking/filtering steps. Check out this work led by @adam_x_yang during his internship with us at @SpotifyResearch w. @enricopalumbo91 and Hugues Bouchard⬇️
Gustavo Penha tweet media
English
1
1
11
376
Gustavo Penha retweetledi
Sumit
Sumit@_reachsumit·
Adaptive Repetition for Mitigating Position Bias in LLM-Based Ranking Spotify introduces a dynamic early-stopping method that adaptively determines repetitions needed for each ranking instance, reducing LLM calls by 81% while preserving accuracy. 📝arxiv.org/abs/2507.17788
English
0
2
10
809
Gustavo Penha retweetledi
Sumit
Sumit@_reachsumit·
Aligned Query Expansion: Efficient Query Expansion for Information Retrieval through LLM Alignment @adam_x_yang et al. leverage LLM alignment techniques to fine-tune models for generating query expansions that directly optimize retrieval effectiveness. 📝arxiv.org/abs/2507.11042
English
0
1
6
524
Gustavo Penha retweetledi
Sumit
Sumit@_reachsumit·
Contextualizing Spotify's Audiobook List Recommendations with Descriptive Shelves Spotify introduces a pipeline that generates personalized audiobook recommendations with descriptive shelves to help users explore content based on their interests. 📝arxiv.org/abs/2504.13572
English
0
4
14
597
Gustavo Penha
Gustavo Penha@_Guz_·
The best-performing ID strategy was to use collaborative-filtering embeddings as input to the discretization approach for semantic IDs
Gustavo Penha tweet media
English
1
0
1
115