Public Policy Analytics papers, 2023–2026

Key findings

• Public Policy Analytics went from 16 papers in 2023 to 29 in 2026, roughly 1.8x in three years.
• NeurIPS publishes the most Public Policy Analytics work: 45 papers from 2023 to 2026, 54% of the four-venue total.
• The fastest-growing overlap is with Reinforcement Learning: papers tagged with both rose from 6 in 2023 to 18 in 2026.
• Among the 5 topics under AI for Social Good, it ranks #2 by 2026 paper count; the fastest grower in the group, Energy Grid Optimization, is up 2.6x versus 1.8x here.

Representative papers

Spread across years and venues so every edition shows up

• NeurIPS 2026Co-Evolving Policy Distillation
• ICML 2026Partial Identification of Policy Values under Network Interference
• ICLR 2026Joint Selection for Large-Scale Pre-Training Data via Policy Gradient-based Mask Learning
• CVPR 2026Real-Time Long Horizon Air Quality Forecasting via Group-Relative Policy Optimization
• NeurIPS 2025Real-Time Hyper-Personalized Generative AI Should Be Regulated to Prevent the Rise of "Digital Heroin"
• ICML 2025Improving Value Estimation Critically Enhances Vanilla Policy Gradient
• CVPR 2025AffordDP: Generalizable Diffusion Policy with Transferable Affordance
• NeurIPS 2024SafeWorld: Geo-Diverse Safety Alignment
• ICML 2024Optimal Differentially Private Model Training with Public Data
• NeurIPS 2023Creating a Public Repository for Joining Private Data
• ICML 2023Exponential Smoothing for Off-Policy Learning
• ICLR 2023Linear Convergence of Natural Policy Gradient Methods with Log-Linear Policies
• NeurIPS 2026Distributionally-Robust Policy Learning from Observational Data
• ICML 2026Personalized Policy Learning through Discrete Experimentation
• NeurIPS 2025$\textit{Hyper-GoalNet}$: Goal-Conditioned Manipulation Policy Learning with HyperNetworks
• ICML 2025Theoretical guarantees on the best-of-n alignment policy
• NeurIPS 2024Improved off-policy training of diffusion samplers
• ICML 2024Position: Considerations for Differentially Private Learning with Large-Scale Public Pretraining
• NeurIPS 2023Global Optimality and Finite Sample Analysis of Softmax Off-Policy Actor Critic under State Distribution Mismatch
• ICML 2023On Many-Actions Policy Gradient

All Public Policy Analytics papers

84 papers, page 1 of 2

1. CVPR 2026Real-Time Long Horizon Air Quality Forecasting via Group-Relative Policy Optimization
2. ICLR 2026Joint Selection for Large-Scale Pre-Training Data via Policy Gradient-based Mask Learning
3. ICML 2026Partial Identification of Policy Values under Network Interference
4. ICML 2026Personalized Policy Learning through Discrete Experimentation
5. ICML 2026Recurrent Structural Policy Gradient for Partially Observable Mean Field Games
6. ICML 2026Reusing Trajectories in Policy Gradients Enables Fast Convergence
7. ICML 2026Unlearning with Asymmetric Sources: Improved Unlearning-Utility Trade-off with Public Data
8. NeurIPS 2026Beyond SFT-to-RL: Pre-alignment via Black-box On-policy Distillation for Multimodal RL
9. NeurIPS 2026Boosting Off-Policy RLVR with Data-Centric Replay
10. NeurIPS 2026Co-Evolving Policy Distillation
11. NeurIPS 2026Compose Your Oracles: Off Policy Improvement with Aggregated Guidance
12. NeurIPS 2026Cornerstones or Stumbling Blocks? Deciphering the Rock Tokens in On-Policy Distillation
13. NeurIPS 2026Distributionally-Robust Policy Learning from Observational Data
14. NeurIPS 2026Don’t Let Gains FADE: Breaking Down Policy Gradient Weights in RL
15. NeurIPS 2026Feedback World Model Enables Precise Guidance of Diffusion Policy
16. NeurIPS 2026FLAG: Flow Policy MaxEnt-RL by Latent Augmented Guidance
17. NeurIPS 2026Global linear convergence of entropy-regularized softmax policy gradient beyond tabular MDPs
18. NeurIPS 2026Jump Start Your Policy Learning with Lessons from 145,000 Training Runs
19. NeurIPS 2026Learning Multimodal One-step Flow Policy via Value-weighted Optimal Transport
20. NeurIPS 2026METAFORGET: Audit-Driven Update-Policy Learning for Reliable Language Model Unlearning
21. NeurIPS 2026Off-policy Learning with Excursion Policies
22. NeurIPS 2026Policy Gradient Algorithms in Average-Reward Multichain MDPs
23. NeurIPS 2026Policy Regret Minimization in Partially Observable Markov Games
24. NeurIPS 2026Proposing Better Rollouts for On-Policy Distillation
25. NeurIPS 2026Reverse to Advance: Teleoperation-Cost Effective Hard Policy Learning from Reversed Easy Tasks
26. NeurIPS 2026Selective Off-Policy Reference Tuning with Plan Guidance
27. NeurIPS 2026The AI Observatory: A Public Measure of Real-World AI Use
28. NeurIPS 2026Trust Region Policy Distillation
29. NeurIPS 2026Your Teacher Can’t Help You Here: Combating Supervision Fidelity Decay in On-Policy Distillation
30. CVPR 2025AffordDP: Generalizable Diffusion Policy with Transferable Affordance
31. ICML 2025Convergence Analysis of Policy Gradient Methods with Dynamic Stochasticity
32. ICML 2025Distributed Differentially Private Data Analytics via Secure Sketching
33. ICML 2025Improving Value Estimation Critically Enhances Vanilla Policy Gradient
34. ICML 2025Return Capping: Sample Efficient CVaR Policy Gradient Optimisation
35. ICML 2025The Value of Prediction in Identifying the Worst-Off
36. ICML 2025Theoretical guarantees on the best-of-n alignment policy
37. NeurIPS 2025$\textit{Hyper-GoalNet}$: Goal-Conditioned Manipulation Policy Learning with HyperNetworks
38. NeurIPS 20253D Equivariant Visuomotor Policy Learning via Spherical Projection
39. NeurIPS 2025Does Stochastic Gradient really succeed for bandits?
40. NeurIPS 2025Flattening Hierarchies with Policy Bootstrapping
41. NeurIPS 2025FreqPolicy: Efficient Flow-based Visuomotor Policy via Frequency Consistency
42. NeurIPS 2025Latent Policy Barrier: Learning Robust Visuomotor Policies by Staying In-Distribution
43. NeurIPS 2025NeurIPS should lead scientific consensus on AI policy
44. NeurIPS 2025Policy Compatible Skill Incremental Learning via Lazy Learning Interface
45. NeurIPS 2025Real-Time Hyper-Personalized Generative AI Should Be Regulated to Prevent the Rise of "Digital Heroin"
46. NeurIPS 2025Regret Analysis of Average-Reward Unichain MDPs via an Actor-Critic Approach
47. NeurIPS 2025Structure Matters: Dynamic Policy Gradient
48. ICML 2024ACE: Off-Policy Actor-Critic with Causality-Aware Entropy Regularization
49. ICML 2024Building Socially-Equitable Public Models
50. ICML 2024Fair Off-Policy Learning from Observational Data

FAQ

How many Public Policy Analytics papers were published at NeurIPS, ICLR, ICML and CVPR from 2023 to 2026?

84 in total: 16 in 2023, 21 in 2024, 18 in 2025, 29 in 2026.

How many Public Policy Analytics papers are at NeurIPS 2026?

NeurIPS 2026 has 22 Public Policy Analytics papers, versus 11 the year before.

How many Public Policy Analytics papers are at ICLR 2026?

ICLR 2026 has 1 Public Policy Analytics papers.

How many Public Policy Analytics papers are at ICML 2026?

ICML 2026 has 5 Public Policy Analytics papers, versus 6 the year before.

How many Public Policy Analytics papers are at CVPR 2026?

CVPR 2026 has 1 Public Policy Analytics papers, versus 1 the year before.

Which conference publishes the most Public Policy Analytics papers?

NeurIPS, with 45 papers from 2023 to 2026.