No article text was recorded for this version.
warm-start policy initialization
Version 2 of 2264 papers305 conceptsfrozen 2026-10-04
Papers in this version
- Blog: How Much Backtracking is Enough? Exploring the Interplay of SFT and RL in Enhancing LLM Reasoning
- Blog: Learning End-to-End Goal-Oriented Dialog with Multiple Answers
- Blog: Mind the Gap: Bridging Thought Leap for Improved Chain-of-Thought Tuning
- Blog: Demystifying Long Chain-of-Thought Reasoning in LLMs
- Blog: WebWatcher: Breaking New Frontiers of Vision-Language Deep Research Agent
- Blog: Balancing the Budget: Understanding Trade-offs Between Supervised and Preference-Based Finetuning
- Blog: ReviewRL: Towards Automated Scientific Review with RL
- Blog: Long Grounded Thoughts: Distilling Compositional Visual Reasoning Chains at Scale
- Blog: Vanishing Gradients in Reinforcement Finetuning of Language Models
- Blog: Improving the Reasoning of Multi-Image Grounding in MLLMs via Reinforcement Learning
- Blog: MedFact-R1: Towards Factual Medical Reasoning via Pseudo-Label Augmentation
- Blog: Toward Effective Reinforcement Learning Fine-Tuning for Medical VQA in Vision-Language Models
- Blog: TinyLLaVA-Video-R1: Towards Smaller LMMs for Video Reasoning
- Blog: Reinforcement Learning on Pre-Training Data
- Blog: Demystifying Long Chain-of-Thought Reasoning in LLMs
- Blog: WebSailor: Navigating Super-human Reasoning for Web Agent
- Blog: One Missing Piece for Open-Source Reasoning Models: A Dataset to Mitigate Cold-Starting Short CoT LLMs in RL
- Blog: GTA: Supervised-Guided Reinforcement Learning for Text Classification with Large Language Models
- Blog: R3-RAG: Learning Step-by-Step Reasoning and Retrieval for LLMs via Reinforcement Learning
- Blog: Through the Valley: Path to Effective Long CoT Training for Small Language Models
- Blog: SFT Memorizes, RL Generalizes: A Comparative Study of Foundation Model Post-training
- Blog: An Actor-Critic Algorithm for Sequence Prediction
- Blog: ThinkDial: An Open Recipe for Controlling Reasoning Effort in Large Language Models
- Blog: Datasets and Recipes for Video Temporal Grounding via Reinforcement Learning
- Blog: A Reinforcement Learning-driven Translation Model for Search-Oriented Conversational Systems
- Blog: Customer-R1: Personalized Simulation of Human Behaviors via RL-based LLM Agent in Online Shopping
- Blog: Balancing the Budget: Understanding Trade-offs Between Supervised and Preference-Based Finetuning
- Blog: UserRL: Training Interactive User-Centric Agent via Reinforcement Learning
- Blog: Vision-R1: Incentivizing Reasoning Capability in Multimodal Large Language Models
- Blog: DeepEyesV2: Toward Agentic Multimodal Model
- Blog: Handover Control in Wireless Systems via Asynchronous Multiuser Deep Reinforcement Learning
- Blog: Fin-R1: A Large Language Model for Financial Reasoning through Reinforcement Learning
- Blog: Enigmata: Scaling Logical Reasoning in Large Language Models with Synthetic Verifiable Puzzles
- Blog: Dual Learning for Machine Translation
- Blog: SARI: Structured Audio Reasoning via Curriculum-Guided Reinforcement Learning
- Blog: DeepResearch-9K: A Challenging Benchmark Dataset of Deep-Research Agent
- Blog: Skin-R1: Toward Trustworthy Clinical Reasoning for Dermatological Diagnosis
- Blog: Beneficial Reasoning Behaviors in Agentic Search and Effective Post-training to Obtain Them
- Blog: Hyperparameter Optimization for Tracking With Continuous Deep Q-Learning
- Blog: Kimina-Prover Preview: Towards Large Formal Reasoning Models with Reinforcement Learning
- Blog: Towards A Better Initial Policy Model For Scalable Long-CoT Reinforcement Learning
- Blog: Artemis: Structured Visual Reasoning for Perception Policy Learning
- Blog: TRANS-ZERO: Self-Play Incentivizes Large Language Models for Multilingual Translation Without Parallel Data
- Blog: Long Grounded Thoughts: Synthesizing Visual Problems and Reasoning Chains at Scale
- Blog: WeFT: Weighted Entropy-driven Fine-Tuning for dLLMs
- Blog: Hierarchical Reinforcement Learning for Integrated Recommendation
- Blog: 3D-R1: Enhancing Reasoning in 3D VLMs for Unified Scene Understanding
- Blog: WebAgent-R1: Training Web Agents via End-to-End Multi-Turn Reinforcement Learning
- Blog: RLHF Deciphered: A Critical Analysis of Reinforcement Learning from Human Feedback for LLMs
- Blog: Consolidating Reinforcement Learning for Multimodal Discrete Diffusion Models
- Blog: Classification with Costly Features using Deep Reinforcement Learning
- Blog: R1-Omni: Explainable Omni-Multimodal Emotion Recognition with Reinforcement Learning
- Blog: ProFit: Leveraging High-Value Signals in SFT via Probability-Guided Token Selection
- Blog: Reinforcement Learning for Self-Improving Agent with Skill Library
- Blog: Self-Improving World Modelling with Latent Actions
- Blog: LLMR: Knowledge Distillation with a Large Language Model-Induced Reward
- Blog: CoDistill-GRPO: A Co-Distillation Recipe for Efficient Group Relative Policy Optimization
- Blog: FIRE: Learning to Navigate and Act on Real-World Files via Stateful Reinforcement Learning
- Blog: Using RL to Identify Divisive Perspectives Improves LLMs Abilities to Identify Communities on Social Media
- Blog: Demystifying Long Chain-of-Thought Reasoning
Showing 60 of 264 papers.
History & citation
Current · Permanent link
Cite this direction
Latest version (a living page):
@misc{lacuna_warm_start_policy_initialization_7e8ba44d,
title = {warm-start policy initialization},
author = {{Lacuna}},
howpublished = {Lacuna research direction},
year = {2026},
url = {https://lacuna.tiptreesystems.com/direction/warm-start-policy-initialization/rdr_7e8ba44dd295460787882b8b8ebd2c10},
note = {Latest version of a living research direction; cite a version for a fixed text. Accessed 2026-11-21},
}This exact version (version 2, frozen text):
@misc{lacuna_warm_start_policy_initialization_7e8ba44d_v2,
title = {warm-start policy initialization},
author = {{Lacuna}},
howpublished = {Lacuna research direction},
year = {2026},
url = {https://lacuna.tiptreesystems.com/direction/warm-start-policy-initialization/rdr_7e8ba44dd295460787882b8b8ebd2c10/v/2},
note = {Version 2, frozen 2026-10-04. Accessed 2026-11-21},
}