Yian Ma
Assistant Professor
Develops scalable Bayesian inference and sampling algorithms with computational and statistical guarantees, focused on time series and sequential decisions.
Yian Ma's lab at UCSD works on reinforcement learning, learning theory & optimization and efficient training & inference. 41 papers since 2023, 36 at top venues, mostly NeurIPS and ICML. 3 alumni, most often to Snowflake AI. Applications close .
At a glance
- papers, 2023–2026
- 11/ year
- at T1 venues
- 88%
- PhD students
- 0
- MS students
- 0
- alumni listed
- 3
Papers by yearfrom the lab homepage, 2023–2026
- 9 papers2026 · T1 8 · collecting
- 11 papers2025 · T1 10
- 13 papers2024 · T1 12
- 8 papers2023 · T1 6
T1 venuesOtherVenues: NeurIPS, ICML, ICLR, AISTATS, COLT
Research focuspaper titles that use each topic's words
- Learning theory & optimization11 papers
- Reinforcement learning7 papers
- Efficient training & inference6 papers
- develops scalable0 papers
- bayesian inference0 papers
Paper list
From the lab homepage.
20269 papers
- COLT (T1)Almost Linear Convergence under Minimal Score Assumptions: Quantized Transition Diffusion (opens in a new tab)
- ICLR (T1)Beyond Length: Quantifying Long-Range Information for Long-Context LLM Pretraining Data (opens in a new tab)
- NeurIPS (T1)Calibrating LLMs with Semantic-level Reward (opens in a new tab)
- ICML (T1)Divide and Learn: Multi-Objective Combinatorial Optimization at Scale (opens in a new tab)
- NeurIPS (T1)Equilibrium Forcing: Adaptive Video Generation Without Noise Conditioning (opens in a new tab)
- ICLR (T1)LaDiR: Latent Diffusion Enhances LLMs for Text Reasoning (opens in a new tab)
- COLMSimulRAG: Simulator-based RAG for Grounding LLMs in Long-form Scientific QA (opens in a new tab)
- ICML (T1)Stochastic Gradient Variational Inference with Price’s Gradient Estimator from Bures-Wasserstein to Parameter Space (opens in a new tab)
- ICLR (T1)Zephyrus: An Agentic Framework for Weather Science (opens in a new tab)
202511 papers
- SIAM Journal on Applied Dynamical SystemsA Gradient-Based Optimization Method Using the Koopman Operator (opens in a new tab)
- ICLR (T1)A Skewness-Based Criterion for Addressing Heteroscedastic Noise in Causal Discovery (opens in a new tab)
- AISTATS (T1)Accuracy on the Wrong Line: On the Pitfalls of Noisy Data for Out-of-distribution Generalisation (opens in a new tab)
- ICLR (T1)ClimaQA: An Automated Evaluation Framework for Climate Foundation Models (opens in a new tab)
- AISTATS (T1)Diffusion Models as Constrained Samplers for Optimization with Unknown Constraints (opens in a new tab)
- ICML (T1)Discovering Latent Structural Causal Models from Spatio-Temporal Data (opens in a new tab)
- NeurIPS (T1)Efficiently Scaling LLM Reasoning Programs with Certaindex (opens in a new tab)
- ICML (T1)Learning to Steer Learners in Games (opens in a new tab)
- NeurIPS (T1)Nearly Dimension-Independent Convergence of Mean-Field Black-Box Variational Inference (opens in a new tab)
- NeurIPS (T1)Optimization on Pareto Sets: On a Theory of Multi-Objective Optimization (opens in a new tab)
- NeurIPS (T1)Purifying Approximate Differential Privacy with Randomized Post-processing (opens in a new tab)
202413 papers
- ICML (T1)Demystifying Doubly Stochastic Gradient Descent (opens in a new tab)
- ICML (T1)Discovering Mixtures of Structural Causal Models from Time Series Data (opens in a new tab)
- Physica D: Nonlinear Phenomena, 460, 134053Estimate exponential memory decay in hidden Markov model and its applications to inference (opens in a new tab)
- ICML (T1)Faster Sampling via Stochastic Gradient Proximal Sampler (opens in a new tab)
- COLT (T1)Faster Sampling without Isoperimetry via Diffusion-based Monte Carlo (opens in a new tab)
- AISTATS (T1)Learning Granger Causality from Instance-wise Self-attentive Hawkes Processes (opens in a new tab)
- AISTATS (T1)Linear Convergence of Black-Box Variational Inference: Should We Stick the Landing? (opens in a new tab)
- NeurIPS (T1)Log-concave Sampling from a Convex Body with a Barrier: A Robust and Unified Dikin Walk (opens in a new tab)
- ICML (T1)Multi-Fidelity Residual Neural Processes for Scalable Surrogate Modeling (opens in a new tab)
- UAI (T1)On Convergence of Federated Averaging Langevin Dynamics (opens in a new tab)
- ICLR (T1)Reverse Diffusion Monte Carlo (opens in a new tab)
- NeurIPS (T1)Reverse Transition Kernel: A Flexible Framework to Accelerate Diffusion Inference (opens in a new tab)
20238 papers
- NeurIPS (T1)Aiming towards the Minimizers: Fast Convergence of SGD for Overparametrized Problems (opens in a new tab)
- Journal of the American Statistical Association (JASA), 0, 1-12Bayesian Robustness: A Nonasymptotic Viewpoint (opens in a new tab)
- KDD (T1)Deep Bayesian Active Learning for Accelerating Stochastic Simulation (opens in a new tab)
- ICML (T1)Disentangled Multi-Fidelity Deep Bayesian Active Learning (opens in a new tab)
- ICML (T1)Langevin Thompson Sampling with Logarithmic Communication: Bandits and Reinforcement Learning (opens in a new tab)
- NeurIPS (T1)On the Convergence of Black-Box Variational Inference (opens in a new tab)
- NeurIPS (T1)Posterior Sampling with Delayed Feedback for Reinforcement Learning with Linear Function Approximation (opens in a new tab)
- SIAM Journal on Applied Dynamical Systems (SIADS), 22 (3), 1523-1551The Adaptive Spectral Koopman Method for Dynamical Systems (opens in a new tab)
Where alumni went3 listed, 50% to industry
Industry 1Academia 1
Most often to
- Snowflake AI
Also to
- Stanford University
Apply to UCSDFall 2027
UCSD CSE Computer Science PhD
- Round
- Single round
- Opens
- —
- GRE
- See program page
- English
- See program page
- Fee
- See program page
Not yet confirmed for this intake.
Similar labsnearest by research summary
- SGNUS
Bryan Kian Hsiang Low
- Probabilistic & causal inference
- USUCSD
Biwei Huang
- Probabilistic & causal inference
- GBUCL
Andrew Saxe
- Computational neuroscience & cognition
- GBImperial
Yassir Jedra
- Reinforcement learning
- CNZhejiang
Haishuai Wang
- Time series, anomaly & spatio-temporal data
- USCMU
Eric Xing
- Generative & foundation models
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.
UCSDcloses