Wang shuai (Dylan)

Postdoc at @Ielabgroup in @UQSchoolEECS | @UWAEMS alumni

RAG, IR, NLP, Model Efficiency

BlueSky: dylanshuaiwang

[Wang shuai (Dylan) @dylan_wangs]

how it works: append K [MASK]s to a retrieval prompt; the diffusion LM fills them in one bidirectional forward pass → K dense (ColBERT MaxSim) + K sparse (max-pooled logits)

joint work with Yin Yu @ShengyaoZhuang @BevanKoopman @GuidoZuccon (UQ + CSIRO)

[Wang shuai (Dylan) @dylan_wangs]

DiffRetriever
Turn a diffusion LM into a retriever: both as dense/sparse/hybrid retriever.
→Works both zero-shot or tuned!
→Beats PromptReps, DiffEmbed, RepLLaMA when tuned
→~15× faster than AR multi-token (zero-shot) and 3x AR multi-token (tuned)
https://t.co/XSfI80vE9R

[Wang shuai (Dylan) @dylan_wangs]

@joelmmackenzie That's Data compression! @joelmmackenzie the role suits me perfectly

[Wang shuai (Dylan) @dylan_wangs]

@joelmmackenzie Hi Dr Makenzie, I'm looking for a PhD at the moment

[Wang shuai (Dylan) @dylan_wangs]

What if you can turn any paper into a live demo with one command?
for example: paper-demo-agent demo 1706.03762

It reads the paper and builds demo end-to-end(app, slides, project page, diagram...).

10 output types · 6 LLM
Picture show example paper

⭐ https://t.co/3RSBsOaen8

[Wang shuai (Dylan) @dylan_wangs]

on top of that, if you also worry only about getting the correct Mesh Terms, you can also use our Mesh-Suggester Tool
🤗 https://t.co/MrlufnJGM6
📷 In WSDM2023 demo paper: https://t.co/p3aGblbyti
The old cloud service clashed so moved it to hugging-face.

[Wang shuai (Dylan) @dylan_wangs]

🚀 AutoBool Demo is live!

An 4B-LLM that automatically generates Boolean queries for systematic reviews: outperforms GPT-4o & O3 on extensive evaluation

🤗 https://t.co/G4C5ESiBep
📄 Accepted in EACL 2026 Main Oral: https://t.co/u7ZZwkJOLU

#SystematicReview #EBM #NLP #Pubmed

[Sumit @_reachsumit]

Beyond Chunk-Then-Embed: A Comprehensive Taxonomy and Evaluation of Document Chunking Strategies for Information Retrieval

Provides a systematic evaluation of doc chunking strategies for dense retrieval, finding that optimal methods are task-dependent.

📝https://t.co/EAoS37Yek5

[Wang shuai (Dylan) @dylan_wangs]

I'm here in Sydney for WWW2025! see us this Thursday, at WWW conference at Posterboard-21, Short Paper & Demo Parkside Ballroom if you are interested!

[Wang shuai (Dylan) @dylan_wangs]

Today in @SIGIRConf at M3.1, I will briefly introduce our newly published collection FeB4RAG! The collection is the first collection to evaluate Federated search in RAG. Our resource support evaluation of Resource Selection, Merging, and Human Evaluation on Answer Generation!

[Shuoqi Sun @shuoqi_sun]

Thank you so much @guidozuc ! It was great to collaborate with you and ieLab members @dylan_wangs @ShengyaoZhuang ! Hope to see you soon and work together again. 😁

[Guido Zuccon @guidozuc]

Yesterday I presented our #ECIR2025 paper "An Investigation of Prompt Variations for Zero-shot LLM-based Rankers", with @shuoqi_sun @ShengyaoZhuang @dylan_wangs. Yes, it was in a church! Slides at https://t.co/mpC8zN0rHO and paper at https://t.co/R0L8S9RfQ5

[ielab @IELabGroup]

🎉 Excited to share the list of our @IELabGroup papers accepted at #SIGIR2025 @ACMSIGIR! Topics include efficiency in IR, domain-specific IR (SysRev, podcast), screenshot retrievers, hallucination detection, and more. See you in Padua 🇮🇹!
Details coming soon, stay tuned!

[Sumit @_reachsumit]

Rank-R1: Enhancing Reasoning in LLM-based Document Rerankers via Reinforcement Learning

@ShengyaoZhuang et al. introduce a LLM-based reranker that performs reasoning over queries and candidate documents before ranking.

📝https://t.co/ptj5wHqfip

[Wang shuai (Dylan) @dylan_wangs]

Share our paper "ReSLLM: Large Language Models are Strong Resource Selectors for Federated Search" with @ShengyaoZhuang , @bevan_koopman and @guidozuc has been accepted to TheWebConference 2025 as a short.
Preprint: https://t.co/jq2vmSPlfZ
Github: https://t.co/tuQUgjxCQu

[Wang shuai (Dylan) @dylan_wangs]

@ShengyaoZhuang @bevan_koopman @guidozuc In this paper, we investigate how to use LLMs to generate Synthetic Labels to tune a downstream resource selectors in Federated Search (An Augmented Search Strategy selecting search engines for user retrieval results)

[Shuoqi Sun @shuoqi_sun]

Delighted to share our full paper "An Investigation of Prompt Variations for Zero-shot LLM-based Rankers" w/@ShengyaoZhuang @dylan_wangs @guidozuc has been accepted at @ecir2025! Huge thanks to co-authors. Check it out if you're into prompt variations & LLM-rankers. #ecir2025

[Hang Li @hanglics]

@OlegZendel

[Damiano Spina 🦋@damianospina.com @damiano10]

Unofficial social mixer at Day 2 of #sigirap2024 #naarm

@guidozuc we’re sending the bill 😝

[Wang shuai (Dylan) @dylan_wangs]

Mark Sanderson @IR_oldie is delivering a keynote for SIGIR-AP @ACMSIGIR_AP Melbourne hub about opportunities and challenges to evaluating generative retrieval.
Mentioned about paper
https://t.co/HuTv9CGOVU