A Closed-Form Upper Bound for Admissible Learning-Rate Steps in Belief-Space Dynamics
Abstract
Admissibility in learning-rate steps is characterized by contractivity in KL/Bregman geometry, providing a formulaic upper bound rather than a tunable parameter.
Learning-rate steps are usually treated as hyperparameters. This paper isolates a local beliefspace calculation: when an update is modeled as a projected forward step on the probability simplex, admissibility means contractivity in the natural KL/Bregman geometry. Under this model, the upper bound of an admissible step is not a tuning slogan but a formula.
Community
Learning-rate steps are usually treated as hyperparameters. This paper isolates a local belief-space calculation: when an update is modeled as a projected forward step on the probability simplex, admissibility means contractivity in the natural KL/Bregman geometry. Under this model, the upper bound of an admissible step is not a tuning slogan but a formula.
This is an automated message from the Librarian Bot. I found the following papers similar to this paper.
The following papers were recommended by the Semantic Scholar API
- Maximizing Rollout Informativeness under a Fixed Budget: A Submodular View of Tree Search for Tool-Use Agentic Reinforcement Learning (2026)
- DDO-RM: Distribution-Level Policy Improvement after Reward Learning (2026)
- A Representation Optimization Dichotomy, Lie-Algebraic Policy Optimization (2026)
- When Policies Cannot Be Retrained: A Unified Closed-Form View of Post-Training Steering in Offline Reinforcement Learning (2026)
- Stabilizing LLM Supervised Fine-Tuning via Explicit Distributional Control (2026)
- Listwise Policy Optimization: Group-based RLVR as Target-Projection on the LLM Response Simplex (2026)
- Global Optimality for Constrained Exploration via Penalty Regularization (2026)
Please give a thumbs up to this comment if you found it helpful!
If you want recommendations for any Paper on Hugging Face checkout this Space
You can directly ask Librarian Bot for paper recommendations by tagging it in a comment: @librarian-bot recommend
Get this paper in your agent:
hf papers read 2605.06741 Don't have the latest CLI?
curl -LsSf https://hf.co/cli/install.sh | bash Models citing this paper 0
No model linking this paper
Datasets citing this paper 0
No dataset linking this paper
Spaces citing this paper 0
No Space linking this paper
Collections including this paper 0
No Collection including this paper