Lin Yang
Associate Professor
Develops provable algorithms for reinforcement learning, LLM acceleration and quantization, and non-convex optimization.
Lin Yang's lab at UCLA works on learning theory & optimization, reinforcement learning and efficient training & inference. 37 papers since 2023, 27 at top venues, mostly NeurIPS and ICLR. 6 alumni. Applications close .
At a glance
- papers, 2023–2026
- 10/ year
- at T1 venues
- 73%
- PhD students
- 0
- MS students
- 0
- alumni listed
- 6
Papers by yearfrom DBLP, 2023–2026
- 4 papers2026 · T1 2 · collecting
- 13 papers2025 · T1 12
- 10 papers2024 · T1 4
- 10 papers2023 · T1 9
T1 venuesOtherVenues: NeurIPS, ICLR, ICML, ISIT, AISTATS
Research focuspaper titles that use each topic's words
- Learning theory & optimization18 papers
- Reinforcement learning18 papers
- Efficient training & inference6 papers
- non-convex optimization0 papers
- non-convex0 papers
Paper list
From DBLP.
20264 papers
- ICLR (T1)ARMOR: High-Performance Semi-Structured Pruning via Adaptive Matrix Factorization (opens in a new tab)
- ISITBest-Arm Identification with Noisy Actuation (opens in a new tab)
- IEEE J. Sel. Areas Inf. TheoryFundamental Limits of Learning Under Erasure-Constrained Communication Channels (opens in a new tab)
- ICLR (T1)Near-Optimal Sample Complexity Bounds for Constrained Average-Reward MDPs (opens in a new tab)
202513 papers
- NeurIPS (T1)Breaking the Frozen Subspace: Importance Sampling for Low-Rank Optimization in LLM Pretraining (opens in a new tab)
- NeurIPS (T1)Combinatorial Ski Rental Problem: Robust and Learning-Augmented Algorithms (opens in a new tab)
- NeurIPS (T1)Deployment Efficient Reward-Free Exploration with Linear Function Approximation (opens in a new tab)
- NeurIPS (T1)Distributed Multi-Agent Bandits Over Erdős-Rényi Random Networks (opens in a new tab)
- ISITDoes Feedback Help in Bandits with Arm Erasures? (opens in a new tab)
- NeurIPS (T1)Federated Multi-armed Bandits with Efficient Bit-Level Communications (opens in a new tab)
- ICML (T1)Hyper: Hyperparameter Robust Efficient Exploration in Reinforcement Learning (opens in a new tab)
- ICLR (T1)Misspecified Q-Learning with Sparse Linear Function Approximation: Tight Bounds on Approximation Error (opens in a new tab)
- NeurIPS (T1)Near-Optimal Sample Complexity for Online Constrained MDPs (opens in a new tab)
- ICLR (T1)Regret-Optimal List Replicable Bandit Learning: Matching Upper and Lower Bounds (opens in a new tab)
- NeurIPS (T1)SD-VLM: Spatial Measuring and Understanding with Depth-Encoded Vision-Language Models (opens in a new tab)
- NeurIPS (T1)Trajectory Graph Learning: Aligning with Long Trajectories in Reinforcement Learning Without Reward Design (opens in a new tab)
202410 papers
- Nat. Mac. Intell.A collective AI via lifelong learning and sharing at the edge (opens in a new tab)
- NeurIPS (T1)Confident Natural Policy Gradient for Local Planning in qπ-realizable Constrained MDPs (opens in a new tab)
- IJCNNDelayed MDPs with Feature Mapping (opens in a new tab)
- IEEE Trans. Wirel. Commun.Federated Learning With Massive Random Access (opens in a new tab)
- AISTATS (T1)Horizon-Free and Instance-Dependent Regret Bounds for Reinforcement Learning with General Function Approximation (opens in a new tab)
- Neural NetworksModeling Bellman-error with logistic distribution with applications in reinforcement learning (opens in a new tab)
- AISTATS (T1)Multi-Agent Bandit Learning through Heterogeneous Action Erasure Channels (opens in a new tab)
- SmartNetsProvably Correct SGD-Based Exploration for Generalized Stochastic Bandit Problem (opens in a new tab)
- NeurIPS (T1)Uniform Last-Iterate Guarantee for Bandits and Reinforcement Learning (opens in a new tab)
- EMBCWearable Body Sensors Integrated into a Virtual Reality Environment - A Modality for Automating the Rehabilitation of the Motor Control System (opens in a new tab)
202310 papers
- COLT (T1)Contexts can be Cheap: Solving Stochastic Contextual Bandits with Linear Bandit Algorithms (opens in a new tab)
- ICML (T1)Distributed Contextual Linear Bandits with Minimax Optimal Communication Cost (opens in a new tab)
- ICML (T1)Does Sparsity Help in Learning Misspecified Linear Bandits? (opens in a new tab)
- NeurIPS (T1)Efficient Batched Algorithm for Contextual Linear Bandits with Large Action Space via Soft Elimination (opens in a new tab)
- ACL (T1)MetaVL: Transferring In-Context Learning Ability From Language Models to Vision-Language Models (opens in a new tab)
- JMLR (T1)Model-Based Multi-Agent RL in Zero-Sum Markov Games with Near-Optimal Sample Complexity (opens in a new tab)
- ISITMulti-Arm Bandits over Action Erasure Channels (opens in a new tab)
- ICLR (T1)Provably Efficient Lifelong Reinforcement Learning with Linear Representation (opens in a new tab)
- NeurIPS (T1)Replicability in Reinforcement Learning (opens in a new tab)
- NeurIPS (T1)Tackling Heavy-Tailed Rewards in Reinforcement Learning with Function Approximation: Minimax Optimal and Instance-Dependent Regret Bounds (opens in a new tab)
Where alumni went6 listed
Apply to UCLAFall 2027
Computer Science PhD
- Round
- Single round (Dec 15)
- Opens
- —
- GRE
- optional
- English
- See program page
- Fee
- See program page
Not yet confirmed for this intake.
Similar labsnearest by research summary
- CNTsinghua
Yong Li
- Time series, anomaly & spatio-temporal data
- CHETH Zurich
Jonas Peters
- Probabilistic & causal inference
- GBCambridge
Sean Holden
- Probabilistic & causal inference
- USCMU
Barnabas Poczos
- Learning theory & optimization
- CHETH Zurich
Bernhard Schölkopf
- Probabilistic & causal inference
- CNZhejiang
Haishuai Wang
- Time series, anomaly & spatio-temporal data
Your notes
Save this lab to keep notes.
Data
Checked against the lab homepage.
- Collected
- Verified
- Papers
- DBLP
- Members
- Lab homepage
- Alumni
- Lab homepage
Tiers are venue classifications, not lab rankings.
UCLAcloses