Robert West
Associate Professor
Heads the Data Science & AI Lab, working at the intersection of natural language processing, AI and computational social science.
Robert West's lab at EPFL works on language & society, signal processing & information theory and computational social science. 69 papers since 2023, 1 at top venues, mostly Science. 5 PhD students. 175 alumni. Applications close .
At a glance
- papers, 2023–2026
- 18/ year
- at T1 venues
- 1%
- PhD students
- 5
- MS students
- 0
- alumni listed
- 175
Papers by yearfrom the lab homepage, 2023–2026
- 14 papers2026 · T1 1 · collecting
- 18 papers2025 · T1 0
- 22 papers2024 · T1 0
- 15 papers2023 · T1 0
T1 venuesOtherVenues: Science
Research focuspaper titles that use each topic's words
- Language & society20 papers
- Computational social science7 papers
- Signal processing & information theory2 papers
- heads0 papers
- social science0 papers
Paper list
From the lab homepage.
202614 papers
- Science (T1)A Narrowing Window to Understand AI (opens in a new tab)
- Nature Human BehaviourA Reporting Checklist for Large Language Models in Behavioural Science (opens in a new tab)
- ICLR WorkshopDiff Mining: Logit Differences Reveal Finetuning Objectives (opens in a new tab)
- CACM --- Communications of the ACMFrom Model Training to Model Raising (opens in a new tab)
- AAAI --- AAAI Conference on Artificial IntelligenceInteractive Evaluation of Large Language Models for Multi-Requirement Software Engineering Tasks (opens in a new tab)
- ICLR --- International Conference on Learning RepresentationsNarrow Finetuning Leaves Clearly Readable Traces in Activation Differences (opens in a new tab)
- Handbook of Computational Social ScienceNavigating Knowledge: Patterns and Insights from Wikipedia Consumption (opens in a new tab)
- ICLR --- International Conference on Learning RepresentationsRL for Reasoning by Adaptively Revealing Rationales (opens in a new tab)
- ICLR --- International Conference on Learning RepresentationsSynthWorlds: Controlled Parallel Worlds for Disentangling Reasoning and Knowledge in Language Models (opens in a new tab)
- EACL --- Conference of the European Chapter of the Association for Computational LinguisticsTandem Training for Language Models (opens in a new tab)
- COLM --- Conference on Language ModelingThe Collaboration Gap: Exploration and Benchmarking of Open-World Agentic Cooperation (opens in a new tab)
- ACL --- Annual Conference of the Association for Computational LinguisticsValid Survey Simulations with Limited Human Data: The Roles of Prompting, Fine-Tuning, and Rectification (opens in a new tab)
202518 papers
- NAACL --- Annual Conference of the North American Chapter of the Association for Computational Linguistics: Human Language TechnologiesA Logical Fallacy-Informed Framework for Argument Generation (opens in a new tab)
- EMNLP --- Conference on Empirical Methods in Natural Language ProcessingCombining Constrained and Unconstrained Decoding via Boosting: BoostCD and Its Application to Information Extraction (opens in a new tab)
- ICLR --- International Conference on Learning RepresentationsControllable Context Sensitivity and the Knob Behind It (opens in a new tab)
- CSCW --- ACM Conference on Computer-Supported Cooperative Work and Social ComputingDeplatforming Norm-Violating Influencers on Social Media Reduces Overall Online Attention Toward Them (opens in a new tab)
- ICWSM --- International AAAI Conference on Web and Social MediaDoes Content Moderation Lead Users Away from Fringe Movements? Evidence from a Recovery Community (opens in a new tab)
- ICML --- International Conference on Machine LearningFleet of Agents: Coordinated Problem Solving with Large Language Models (opens in a new tab)
- EMNLP --- Conference on Empirical Methods in Natural Language ProcessingGRAD: Generative Retrieval-Aligned Demonstration Sampler for Efficient Few-Shot Reasoning (opens in a new tab)
- Findings of ACL --- Annual Conference of the Association for Computational LinguisticsLocalized Cultural Knowledge is Conserved and Controllable in Large Language Models (opens in a new tab)
- Nature Human BehaviourOn the Conversational Persuasiveness of GPT-4 (opens in a new tab)
- NeurIPS --- Neural Information Processing SystemsOne-Step Is Enough: Sparse Autoencoders for Text-to-Image Diffusion Models (opens in a new tab)
- CSCW --- ACM Conference on Computer-Supported Cooperative Work and Social ComputingPost Guidance for Online Communities (opens in a new tab)
- CACM --- Communications of the ACMPrevalence and Prevention of Large Language Model Use in Crowd Work (opens in a new tab)
202422 papers
- ACL --- Annual Conference of the Association for Computational LinguisticsA Glitch in the Matrix? Locating and Detecting Language Model Grounding with Fakepedia (opens in a new tab)
- Findings of EMNLP --- Conference on Empirical Methods in Natural Language ProcessingActivation Scaling for Steering and Interpreting Language Models (opens in a new tab)
- NeurIPS WorkshopAssessing Social Alignment: Do Personality-Prompted Large Language Models Behave Like Humans? (opens in a new tab)
- Findings of EMNLP --- Conference on Empirical Methods in Natural Language ProcessingCan Language Models Recognize Convincing Arguments? (opens in a new tab)
- PNAS --- Proceedings of the National Academy of Sciences of the United States of AmericaCausally Estimating the Effect of YouTube's Recommender System Using Counterfactual Bots (opens in a new tab)
- ICWSM --- International AAAI Conference on Web and Social MediaCurious Rhythms: Temporal Regularities of Wikipedia Consumption (opens in a new tab)
- ACL --- Annual Conference of the Association for Computational LinguisticsDo Llamas Work in English? On the Latent Language of Multilingual Transformers (opens in a new tab)
- EMNLP --- Conference on Empirical Methods in Natural Language ProcessingEntity Insertion in Multilingual Linked Corpora: The Case of Wikipedia (opens in a new tab)
- ICLR --- International Conference on Learning RepresentationsEvaluating Language Model Agency through Negotiations (opens in a new tab)
- WSDM --- International ACM Conference on Web Search and Data MiningFly-Swat or Cannon? Cost-Effective Language Model Choice via Meta-Modeling (opens in a new tab)
- PNAS NexusFood Choice Mimicry on a Large University Campus (opens in a new tab)
- ACL --- Annual Conference of the Association for Computational LinguisticsGetting Serious about Humor: Crafting Humor Datasets with Unfunny Large Language Models (opens in a new tab)
202315 papers
- TWEB --- ACM Transactions on the WebA Large-Scale Characterization of How Readers Browse Wikipedia (opens in a new tab)
- ACL --- Annual Conference of the Association for Computational LinguisticsAn Ordinal Latent Variable Model of Conflict Intensity (opens in a new tab)
- WWW --- The Web ConferenceAutomated Content Moderation Increases Adherence to Community Guidelines (opens in a new tab)
- PNAS NexusDeplatforming Did Not Decrease Parler Users' Activity on Fringe Social Media (opens in a new tab)
- WWW --- The Web ConferenceDescartes: Generating Short Descriptions of Wikipedia Articles (opens in a new tab)
- ICWSM --- International AAAI Conference on Web and Social MediaDiPPS: Differentially Private Propensity Scores for Bias Correction (opens in a new tab)
- SaTML --- IEEE Conference on Secure and Trustworthy Machine LearningDistribution Inference Risks: Identifying and Mitigating Sources of Leakage (opens in a new tab)
- EMNLP --- Conference on Empirical Methods in Natural Language ProcessingExploiting Asymmetry for Synthetic Training Data Generation: SynthIE and the Case of Information Extraction (opens in a new tab)
- EMNLP --- Conference on Empirical Methods in Natural Language ProcessingGrammar-Constrained Decoding for Structured NLP Tasks without Finetuning (opens in a new tab)
- Findings of EACL --- Conference of the European Chapter of the Association for Computational LinguisticsLanguage Model Decoding as Likelihood--Utility Alignment (opens in a new tab)
- PLOS ONELinguistic Effects on News Headline Success: Evidence from Thousands of Online Field Experiments (Registered Report) (opens in a new tab)
- ICWSM --- International AAAI Conference on Web and Social MediaQuotatives Indicate Decline in Objectivity in U.S. Political News (opens in a new tab)
Where alumni went175 listed
Apply to EPFLFall 2027
EDIC Computer and Communication Sciences PhD
- Round
- Round 1
- Opens
- —
- GRE
- optional
- English
- TOEFL 100 / IELTS 7.0
- Fee
- none
Confirmed on the program page.
Similar labsnearest by research summary
- SGNUS
Min-Yen Kan
- Knowledge, QA & retrieval
- USCMU
Fernando Diaz
- Interpretability, evaluation & safety of LMs
- USNYU
Robert Krueger
- Human-AI collaboration & trust
- USMichigan
David Jurgens
- Language & society
- USMichigan
V.G. Vinod Vydiswaran
- Language & society
- CNSJTU
Yaohui Jin
- Large language models
Your notes
Save this lab to keep notes.
Data
Some numbers could not be verified:
- The lab site and the paper record do not agree.
- Collected
- Verified
- Papers
- The lab homepage
- Members
- Lab homepage
- Alumni
- Lab homepage
Tiers are venue classifications, not lab rankings.
EPFLcloses