SM
Sewon Min
cs.CLcs.AIcs.LGcs.IRcs.CVcs.DBcs.NE
On Valency
published · living versionsW_tse9bvhc·v1 · currentpublished
Preprints & journals
67 papers in the corpus · 2016–2026Train An Imaginative Agent2610.26845v1 · Jiaxin Ge, Junyi Zhang, Tianning Chai et al.2026 · 0 citationsarXivon Valency
Scaling Repeated-Data Pretraining: Batch Size, Regularization, and GPU-Time Tradeoffs2610.15965v1 · Sewon Min2026 · 0 citationsarXiv
Base Models Can Reason By Taking a Cue From Training Data2610.06851v1 · Sophie L. Wang, Amil Dravid, Rulin Shao et al.2026 · 0 citationsarXiv
OmniTaskonomy: When Does Visual Generation Improve Visual Understanding?2609.38079v1 · Jiaxin Ge, Yiming Qin, Ji Xie et al.2026 · 0 citationsarXiv
No More Free Lunch: Corpus Task Complexity Matters as Corpora Grow2609.29245v1 · Prasann Singhal, Amanda Bertsch, Jacob Steinhardt et al.2026 · 0 citationsarXiv
Can Language Models Actually Retrieve In-Context? Drowning in Documents at Million Token Scale2607.01538v2 · Siddharth Gollapudi, Nilesh Gupta, Prasann Singhal et al.2026 · 0 citationsarXiv
Reference-Based Distillation Detection in LLMs2607.09692v1 · Rajat Rawat, Sizhe Chen, Akshay Anand et al.2026 · 0 citationsarXiv
Which Models Are Our Models Built On? Auditing Invisible Dependencies in Modern LLMs2606.12385v1 · Sanjay Adhikesaven, Haoxiang Sun, Sewon Min2026 · 0 citationsarXiv
PIXELRAG: Web Screenshots Beat Text for Retrieval-Augmented Generation2606.28344v1 · Yichuan Wang, Zhifei Li, Zirui Wang et al.2026 · 0 citationsarXiv
EMO: Pretraining Mixture of Experts for Emergent Modularity2605.06663v2 · Ryan Wang, Akshita Bhagia, Sewon Min2026 · 0 citationsarXiv
RAG over Thinking Traces Can Improve Reasoning Tasks2605.03344v3 · Negar Arabzadeh, Wenjie Ma, Sewon Min et al.2026 · 0 citationsarXiv
Train Separately, Merge Together: Modular Post-Training with Mixture-of-Experts2604.18473v2 · Jacob Morrison, Sanjay Adhikesaven, Akshita Bhagia et al.2026 · 0 citationsarXiv
Learning to Detect Language Model Training Data via Active Reconstruction2602.19020v1 · Junjie Oscar Yin, John X. Morris, Vitaly Shmatikov et al.2026 · 0 citationsarXiv
Residual Context Diffusion Language Models2601.22954v2 · Yuezhou Hu, Harman Singh, Monishwaran Maheswaran et al.2026 · 0 citationsarXiv
VisGym: Diverse, Customizable, Scalable Environments for Multimodal Agents2601.16973v1 · Zirui Wang, Junyi Zhang, Jiaxin Ge et al.2026 · 0 citationsarXiv
DS SERVE: A Framework for Efficient and Scalable Neural Retrieval2602.22224v1 · Jinjian Liu, Yichuan Wang, Xinxi Lyu et al.2025 · 0 citationsarXiv
RLAC: Reinforcement Learning with Adversarial Critic for Free-Form Generation Tasks2511.01758v1 · Mian Wu, Gavin Zhang, Sewon Min et al.2025 · 0 citationsarXiv
What's In My Human Feedback? Learning Interpretable Descriptions of Preference Data2510.26202v2 · Rajiv Movva, Smitha Milli, Sewon Min et al.2025 · 0 citationsarXiv
Constantly Improving Image Models Need Constantly Improving Benchmarks2510.15021v1 · Jiaxin Ge, Grace Luo, Heekyung Lee et al.2025 · 0 citationsarXiv
Reliable Fine-Grained Evaluation of Natural Language Math Proofs2510.13888v2 · Wenjie Ma, Andrei Cojocaru, Neel Kolhe et al.2025 · 0 citationsarXiv
FlexOlmo: Open Language Models for Flexible Data Use2507.07024v4 · Weijia Shi, Akshita Bhagia, Kevin Farhat et al.2025 · 1 citationarXiv
Frustratingly Simple Retrieval Improves Challenging, Reasoning-Intensive Benchmarks2507.01297v2 · Xinxi Lyu, Michael Duan, Rulin Shao et al.2025 · 0 citationsarXiv
Spurious Rewards: Rethinking Training Signals in RLVR2506.10947v2 · Rulin Shao, Shuyue Stella Li, Rui Xin et al.2025 · 1 citationarXiv
LEANN: A Low-Storage Vector Index2506.08276v2 · Yichuan Wang, Zhifei Li, Shu Liu et al.2025 · 0 citationsarXiv
ReasonIR: Training Retrievers for Reasoning Tasks2504.20595v1 · Rulin Shao, Rui Qiao, Varsha Kishore et al.2025 · 2 citationsarXiv
Reasoning Models Can Be Effective Without Thinking2504.09858v1 · Wenjie Ma, Jingxuan He, Charlie Snell et al.2025 · 0 citationsarXiv
OLMoTrace: Tracing Language Model Outputs Back to Trillions of Training Tokens2504.07096v2 · Jiacheng Liu, Taylor Blanton, Yanai Elazar et al.2025 · 6 citationsarXiv
Organize the Web: Constructing Domains Enhances Pre-Training Data Curation2502.10341v3 · Alexander Wettig, Kyle Lo, Sewon Min et al.2025 · 0 citationsarXiv
OLMoE: Open Mixture-of-Experts Language Models2409.02060v2 · Niklas Muennighoff, Luca Soldaini, Dirk Groeneveld et al.2024 · 8 citationsarXiv
CopyBench: Measuring Literal and Non-Literal Reproduction of Copyright-Protected Text in Language Model Generation2407.07087v2 · Tong Chen, Akari Asai, Niloofar Mireshghallah et al.2024 · 4 citationsarXiv
Scaling Retrieval-Based Language Models with a Trillion-Token Datastore2407.12854v1 · Rulin Shao, Jacqueline He, Akari Asai et al.2024 · 8 citationsarXiv
Do Membership Inference Attacks Work on Large Language Models?2402.07841v2 · Michael Duan, Anshuman Suri, Niloofar Mireshghallah et al.2024 · 17 citationsarXiv
Infini-gram: Scaling Unbounded n-gram Language Models to a Trillion Tokens2401.17377v4 · Jiacheng Liu, Sewon Min, Luke Zettlemoyer et al.2024 · 3 citationsarXiv
In-context Pretraining: Language Modeling Beyond Document Boundaries2310.10638v6 · Weijia Shi, Sewon Min, Maria Lomeli et al.2023 · 3 citationsarXiv
BTR: Binary Token Representations for Efficient Retrieval Augmented Language Models2310.01329v2 · Qingqing Cao, Sewon Min, Yizhong Wang et al.2023 · 1 citationarXiv
SILO Language Models: Isolating Legal Risk In a Nonparametric Datastore2308.04430v2 · Sewon Min, Suchin Gururangan, Eric Wallace et al.2023 · 11 citationsarXiv
FActScore: Fine-grained Atomic Evaluation of Factual Precision in Long Form Text Generation2305.14251v2 · Sewon Min, Kalpesh Krishna, Xinxi Lyu et al.2023 · 414 citationsarXiv
REPLUG: Retrieval-Augmented Black-Box Language Models2301.12652v4 · Weijia Shi, Sewon Min, Michihiro Yasunaga et al.2023 · 274 citationsarXiv
Towards Understanding Chain-of-Thought Prompting: An Empirical Study of What Matters2212.10001v2 · Boshi Wang, Sewon Min, Xiang Deng et al.2022 · 128 citationsarXiv
Z-ICL: Zero-Shot In-Context Learning with Pseudo-Demonstrations2212.09865v2 · Xinxi Lyu, Sewon Min, Iz Beltagy et al.2022 · 24 citationsarXiv
Nonparametric Masked Language Modeling2212.01349v2 · Sewon Min, Weijia Shi, Mike Lewis et al.2022 · 25 citationsarXiv
CREPE: Open-Domain Question Answering with False Presuppositions2211.17257v1 · Xinyan Velocity Yu, Sewon Min, Luke Zettlemoyer et al.2022 · 14 citationsarXiv
Exploring The Landscape of Distributional Robustness for Question Answering Models2210.12517v1 · Anas Awadalla, Mitchell Wortsman, Gabriel Ilharco et al.2022 · 6 citationsarXiv
Measuring and Narrowing the Compositionality Gap in Language Models2210.03350v3 · Ofir Press, Muru Zhang, Sewon Min et al.2022 · 293 citationsarXiv
INSCIT: Information-Seeking Conversations with Mixed-Initiative Interactions2207.00746v2 · Zeqiu Wu, Ryu Parish, Hao Cheng et al.2022 · 12 citationsarXiv
Re-Examining Calibration: The Case of Question Answering2205.12507v2 · Chenglei Si, Chen Zhao, Sewon Min et al.2022 · 11 citationsarXiv
Rethinking the Role of Demonstrations: What Makes In-Context Learning Work?2202.12837v2 · Sewon Min, Xinxi Lyu, Ari Holtzman et al.2022 · 747 citationsarXiv
Prompt Waywardness: The Curious Case of Discretized Interpretation of Continuous Prompts2112.08348v2 · Daniel Khashabi, Shane Lyu, Sewon Min et al.2021 · 27 citationsarXiv
MetaICL: Learning to Learn In Context2110.15943v2 · Sewon Min, Mike Lewis, Luke Zettlemoyer et al.2021 · 220 citationsarXiv
Noisy Channel Language Model Prompting for Few-Shot Text Classification2108.04106v3 · Sewon Min, Mike Lewis, Hannaneh Hajishirzi et al.2021 · 132 citationsarXiv
Career total: 107 works. 67 are in this corpus.Showing the 50 most recent.
Profile built from the corpus for this byline.
Author records are still filling in while the Hub is in alpha. If this is your page, you'll be able to claim it soon. Spot a mistake? Tell us.