Aarti Singh
Professor
Studies statistical machine learning, decision making under uncertainty and interactive learning.
Aarti Singh's lab at CMU works on reinforcement learning and learning theory & optimization. 28 papers since 2023, 15 at top venues, mostly ICML and NeurIPS. 7 PhD students and 1 master's student. 41 alumni, most often to Google and NVIDIA. Applications close .
At a glance
- papers, 2023–2026
- 7/ year
- at T1 venues
- 54%
- PhD students
- 7
- MS students
- 1
- alumni listed
- 41
Papers by yearfrom the lab homepage, 2023–2026
- 10 papers2026 · T1 4 · collecting
- 8 papers2025 · T1 5
- 6 papers2024 · T1 3
- 4 papers2023 · T1 3
T1 venuesOtherVenues: ICML, NeurIPS, AISTATS, ICLR
Research focuspaper titles that use each topic's words
- Reinforcement learning8 papers
- Learning theory & optimization6 papers
- interactive learning0 papers
- studies statistical0 papers
- statistical machine0 papers
Paper list
From the lab homepage.
202610 papers
- ICML (T1)Expanding the Capabilities of Reinforcement Learning via Text Feedback (opens in a new tab)
- The National Academies PressFrontiers of Statistics in Science and Engineering: 2035 and Beyond (opens in a new tab)
- ACM Conference on Fairness, Accountability, and Transparency, FAccT'26Investigating the Role of AI in Emergency Management: Use Cases, Challenges, and Opportunities (opens in a new tab)
- Artifical Intelligence and Statistics, AISTATS'26OEUVRE: OnlinE Unbiased Variance-Reduced loss Estimation (opens in a new tab)
- ICML (T1)On the Learning Dynamics of RLVR at the Edge of Competence (opens in a new tab)
- ICLR (T1)Online Decision Making with Generative Action Sets (opens in a new tab)
- ICML (T1)Online Social Welfare Function-based Resource Allocation (opens in a new tab)
- ICLR WorkshopPhysics-Informed Conditional Diffusion for Multi-Modal PDEs (opens in a new tab)
- Senior Member Track, Annual AAAI Conference on Artificial Intelligence, AAAI'26Scaling Up AI Alignment (opens in a new tab)
- Proceedings of National Academy of Sciences (PNAS) NexusToward a Science of Human-AI Teaming for Decision-making: A complementarity framework (opens in a new tab)
20258 papers
- Transactions on Machine Learning Research (TMLR), JuneBeyond Parameter Count: Implicit Bias in Soft Mixture of Experts (opens in a new tab)
- ICML (T1)Data-driven Design of Randomized Control Trials with Guaranteed Treatment Effects (opens in a new tab)
- PNAS NexusInsectNet: Real-time identification of insects using an end-to-end machine learning pipeline (opens in a new tab)
- AISTATS (T1)Logarithmic Neyman Regret for Adaptive Estimation of the Average Treatment Effect (opens in a new tab)
- ICML (T1)Optimistic Algorithms for Adaptive Estimation of the Average Treatment Effect (opens in a new tab)
- ICML (T1)Projection Optimization: A General Framework for Multi-Objective and Multi-Group RLHF (opens in a new tab)
- Neural Information Processing Systems, NeurIPS'25To Distill or Decide? Understanding the Algorithmic Trade-off in Partially Observable RL (opens in a new tab)
- NeurIPS (T1)Transformers Provably Learn Chain-of-Thought Reasoning with Length Generalization
20246 papers
- Findings of EACLGoodhart's Law Applies to NLP's Explanation Benchmarks (opens in a new tab)
- ICML (T1)Hybrid Reinforcement Learning from Offline Observation Alone (opens in a new tab)
- NeurIPS (T1)Learning Social Welfare Functions (opens in a new tab)
- International Conference on Learning Representations, ICLR'24Role of Locality and Weight Sharing in Image-Based Tasks: A Sample Complexity Separation between CNNs, LCNs, and FCNs (opens in a new tab)
- Journal of Optimization Theory and ApplicationsSpecifying and Solving Robust Empirical Risk Minimization Problems Using CVXPY (opens in a new tab)
- NeurIPS (T1)The Importance of Online Data: Understanding Preference Fine-tuning via Coverage (opens in a new tab)
20234 papers
- AISTATS (T1)Adaptation to Misspecified Kernel Regularization in Kernelised Bandits (opens in a new tab)
- NeurIPS WorkshopPredicting the Initial Conditions of the Universe using a Deterministic Neural Network (opens in a new tab)
- ICML (T1)Virtues of Laziness in Model-based RL: A Unified Objective and Algorithms (opens in a new tab)
- ICML (T1)Weighted Tallying Bandits: Overcoming Intractability via Repeated Exposure Optimality (opens in a new tab)
Where alumni went41 listed, 31% to industry
Industry 8Academia 18
Most often to
- NVIDIA
- Adobe
Also to
- Carnegie Mellon University
- Harvard University
- New Economic School
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
- USUIUC
R. Srikant
- Reinforcement learning
- CNSJTU
Yunbo Wang
- Reinforcement learning
- USUC Berkeley
David Wagner
- Trustworthy & safe ML
- USCMU
Geoffrey J. Gordon
- Learning theory & optimization
- USUC Berkeley
Anil Aswani
- Learning theory & optimization
- CNTsinghua
Xiaolin Hu
- Trustworthy & safe ML
Your notes
Save this lab to keep notes.
Data
Checked against the lab homepage.
- Collected
- Verified
- Papers
- The lab homepage
- Members
- Lab homepage
- Alumni
- Lab homepage
Tiers are venue classifications, not lab rankings.
CMUcloses