No article text was recorded for this version.
forward KL regularization
Version 1 of 124 papers38 conceptsfrozen 2026-10-04
Papers in this version
- Blog: Understanding the Relation Between Maximum-Entropy Inverse Reinforcement Learning and Behaviour Cloning
- Blog: Amortized Population Gibbs Samplers with Neural Sufficient Statistics
- Blog: Bootstrap Off-policy with World Model
- Blog: Diffusion Alignment as Variational Expectation-Maximization
- Blog: ARM: Alignment with Residual Energy-Based Model
- Blog: Low-probability Tokens Sustain Exploration in Reinforcement Learning with Verifiable Reward
- Blog: Accelerating Metropolis-Hastings with Lightweight Inference Compilation
- Blog: Beyond Mode Collapse: Distribution Matching for Diverse Reasoning
- Blog: Sample-Efficient On-Policy Imitation Learning from Observations
- Blog: Reconstructing the Universe with Variational self-Boosted Sampling
- Blog: Anchored Supervised Fine-Tuning
- Blog: Mimicking Better by Matching the Approximate Action Distribution
- Blog: Iterative Distillation for Reward-Guided Fine-Tuning of Diffusion Models in Biomolecular Design
- Blog: Efficient Inference for Coupled Hidden Markov Models in Continuous Time and Discrete Space
- Blog: Token-level Direct Preference Optimization
- Blog: A Divergence Minimization Perspective on Imitation Learning Methods
- Blog: Conditional Image Generation by Conditioning Variational Auto-Encoders
- Blog: disco: a toolkit for Distributional Control of Generative Models
- Blog: On the Correspondence between Compositionality and Imitation in Emergent Neural Communication
- Blog: Mitigating Object Hallucination in MLLMs via Data-augmented Phrase-level Alignment
- Blog: Enhancing Reasoning for Diffusion LLMs via Distribution Matching Policy Optimization
- Blog: Revisiting Generative Policies: A Simpler Reinforcement Learning Algorithmic Perspective
- Blog: KL-Regularized RLHF with Multiple Reference Models: Exact Solutions and Sample Complexity
- Blog: How to model Human Actions distribution with Event Sequence Data
History & citation
Current · Permanent link
- This direction split from Distribution Matching for Diversity in Generative Alignment (54% overlap)
Cite this direction
Latest version (a living page):
@misc{lacuna_forward_kl_regularization_54caae03,
title = {forward KL regularization},
author = {{Lacuna}},
howpublished = {Lacuna research direction},
year = {2026},
url = {https://lacuna.tiptreesystems.com/direction/forward-kl-regularization/rdr_54caae03f7f345b5bf1d710ea4c3217d},
note = {Latest version of a living research direction; cite a version for a fixed text. Accessed 2026-11-22},
}This exact version (version 1, frozen text):
@misc{lacuna_forward_kl_regularization_54caae03_v1,
title = {forward KL regularization},
author = {{Lacuna}},
howpublished = {Lacuna research direction},
year = {2026},
url = {https://lacuna.tiptreesystems.com/direction/forward-kl-regularization/rdr_54caae03f7f345b5bf1d710ea4c3217d/v/1},
note = {Version 1, frozen 2026-10-04. Accessed 2026-11-22},
}