Graham Neubig
Associate Professor
Studies language models and AI agents, including code generation, multilingual NLP and evaluation.
Graham Neubig's lab at CMU works on large language models and reasoning & agents. 124 papers since 2023, 81 at top venues, mostly ICLR and ACL. 14 PhD students and 1 master's student. 37 alumni, most often to IBM. Applications close .
At a glance
- papers, 2023–2026
- 32/ year
- at T1 venues
- 65%
- PhD students
- 14
- MS students
- 1
- alumni listed
- 37
Papers by yearfrom DBLP, 2023–2026
- 21 papers2026 · T1 13 · collecting
- 39 papers2025 · T1 30
- 30 papers2024 · T1 20
- 34 papers2023 · T1 18
T1 venuesOtherVenues: ICLR, ACL, EMNLP, NAACL, ICML
Research focuspaper titles that use each topic's words
- Large language models49 papers
- Reasoning & agents19 papers
- code generation4 papers
- multilingual nlp0 papers
- studies language0 papers
Paper list
From DBLP.
202621 papers
- ICLR (T1)Agent Data Protocol: Unifying Datasets for Diverse, Effective Fine-tuning of LLM Agents (opens in a new tab)
- ICLR (T1)Ambig-SWE: Interactive Agents to Overcome Underspecificity in Software Engineering (opens in a new tab)
- EACLCAIRE: Cultural Attribution of Images with Retrieval (opens in a new tab)
- CHI (T1)Code with Me or for Me? How Increasing AI Automation Transforms Developer Workflows (opens in a new tab)
- Findings of EACLCoding Agents with Multimodal Browsing are Generalist Problem Solvers (opens in a new tab)
- ICLR (T1)EditBench: Evaluating LLM Abilities to Perform Real-World Instructed Code Edits (opens in a new tab)
- ICPRFieldWorkArena: Agentic AI Benchmark for Real Field Work Tasks (opens in a new tab)
- ACL (T1)Gained in Translation: Privileged Pairwise Judges Enhance Multilingual Reasoning (opens in a new tab)
- ICLR (T1)Go-Browse: Training Web Agents with Structured Exploration (opens in a new tab)
- IJCAI/ECAILow-Latency Real-Time Audio Game Commentary System via LLM-Based Parallel Text Generation (opens in a new tab)
- ACL (T1)Massively Multilingual Joint Segmentation and Glossing (opens in a new tab)
- Trans. Assoc. Comput. LinguisticsMERLIN: A Testbed for Multilingual Multimodal Entity Recognition and Linking (opens in a new tab)
202539 papers
- Trans. Mach. Learn. Res.A Gold Standard Dataset for the Reviewer Assignment Problem (opens in a new tab)
- ICML (T1)Agent Workflow Memory (opens in a new tab)
- EMNLP (T1)AgentDiagnose: An Open Toolkit for Diagnosing LLM Agent Trajectories (opens in a new tab)
- MICRO (T1)Automated High-Level Code Optimization for Warehouse Performance (opens in a new tab)
- CVPR (T1)AutoPresent: Designing Structured Visuals from Scratch (opens in a new tab)
- ACL (T1)BehaviorBox: Automated Discovery of Fine-Grained Performance Differences Between Language Models (opens in a new tab)
- NAACL (T1)Benchmarking Failures in Tool-Augmented Language Models (opens in a new tab)
- ICLR (T1)Better Instruction-Following Through Minimum Bayes Risk (opens in a new tab)
- Findings of ACLBeyond Browsing: API-Based Web Agents (opens in a new tab)
- NeurIPS (T1)Checklists Are Better Than Reward Models For Aligning Language Models (opens in a new tab)
- Findings of NAACLCodeRAG-Bench: Can Retrieval Augment Code Generation? (opens in a new tab)
- NAACL (T1)CowPilot: A Framework for Autonomous and Human-Agent Collaborative Web Navigation (opens in a new tab)
202430 papers
- NeurIPS (T1)Alignment for Honesty (opens in a new tab)
- EMNLP (T1)An image speaks a thousand words, but can everyone listen? On image transcreation for cultural relevance (opens in a new tab)
- Findings of ACLBetter Synthetic Data by Retrieving and Transforming Existing Datasets (opens in a new tab)
- IWSLTCMU's IWSLT 2024 Offline Speech Translation System: A Cascaded Approach For Long-Form Robustness (opens in a new tab)
- NAACL (T1)DeMuX: Data-efficient Multilingual Learning (opens in a new tab)
- NeurIPS (T1)Divergences between Language Models and Human Brains (opens in a new tab)
- Trans. Assoc. Comput. LinguisticsDo LLMs Exhibit Human-like Response Biases? A Case Study in Survey Design (opens in a new tab)
- CVPR WorkshopEvaluating and Improving Compositional Text-to-Visual Generation (opens in a new tab)
- ECCV (T1)Evaluating Text-to-Visual Generation with Image-to-Text Generation (opens in a new tab)
- Trans. Mach. Learn. Res.From Decoding to Meta-Generation: Inference-time Algorithms for Large Language Models (opens in a new tab)
- EMNLP (T1)GlossLM: A Massively Multilingual Corpus and Pretrained Model for Interlinear Glossed Text (opens in a new tab)
- IEEE Big DataHILITE: Human-in-the-loop Interactive Tool for Image Editing (opens in a new tab)
202334 papers
- Findings of EACLA Multi-dimensional Evaluation of Tokenizer-free Multilingual Pretrained Models (opens in a new tab)
- ICLR (T1)AANG : Automating Auxiliary Learning (opens in a new tab)
- EMNLP (T1)Active Retrieval Augmented Generation (opens in a new tab)
- ACL (T1)Beyond Contrastive Learning: A Variational Generative Model for Multilingual Retrieval (opens in a new tab)
- Trans. Assoc. Comput. LinguisticsBridging the Gap: A Survey on Integrating (Human) Feedback for Natural Language Generation (opens in a new tab)
- WMTChatGPT MT: Competitive for High- (but Not Low-) Resource Languages (opens in a new tab)
- EMNLP (T1)CodeBERTScore: Evaluating Code Generation with Pretrained Models of Code (opens in a new tab)
- ICLR (T1)Computational Language Acquisition with Theory of Mind (opens in a new tab)
- ICML (T1)Cross-Modal Fine-Tuning: Align then Refine (opens in a new tab)
- EMNLP (T1)Crossing the Threshold: Idiomatic Machine Translation through Retrieval Augmentation and Loss Weighting (opens in a new tab)
- EACLCTC Alignments Improve Autoregressive Translation (opens in a new tab)
- ACL (T1)DataFinder: Scientific Dataset Recommendation from Natural Language Descriptions (opens in a new tab)
Where alumni went37 listed, 50% to industry
Industry 15Academia 15
Most often to
- IBM
Also to
- Carnegie Mellon University
- Mila
- Princeton University
Apply to CMUFall 2027
CMU SCS Computer Science PhD
- Round
- Single round
- Opens
- GRE
- See program page
- English
- TOEFL / IELTS / Duolingo (F-1/J-1 non-native speakers)
- Fee
- USD 100 (USD 80 by early deadline)
Confirmed on the program page.
Similar labsnearest by research summary
- USStanford
Daniel E. Ho
- Large language models
- GBCambridge
Ann Copestake
- Dialogue, generation & summarization
- USUSC
Jonathan May
- Multilingual, translation & linguistics
- USUMD
Mohit Iyyer
- Large language models
- CNSJTU
Pan Hu
- Large language models
- USCMU
Eric Nyberg
- Reasoning & agents
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
- DBLP
- Members
- Lab homepage
- Alumni
- Lab homepage
Tiers are venue classifications, not lab rankings.
CMUcloses